CN102971045B

CN102971045B - Method and apparatus for detecting and discriminating tachycardia

Info

Publication number: CN102971045B
Application number: CN201180032103.0A
Authority: CN
Inventors: X·张; R·W·斯塔德勒; J·M·吉尔伯格
Original assignee: Medtronic Inc
Current assignee: Medtronic Inc
Priority date: 2010-04-28
Filing date: 2011-04-12
Publication date: 2015-07-01
Anticipated expiration: 2031-04-12
Also published as: CN102971045A; WO2011136916A1; EP2563472A1

Abstract

A medical device and associated method for discriminating cardiac events includes sensing a cardiac signal spatially located across approximately a full duration of a predetermined sensing window. A match score is determined corresponding to the sensed cardiac signal. A beat feature of multiple beat features across less than the full duration of the sensing window is determined, the beat feature being selected from the multiple beat features in response to the match score. Cardiac event evidence is accumulated in response to the match score and the determined beat feature, and cardiac events are discriminated in response to the accumulated cardiac evidence.

Description

Detect and distinguish tachycardic method and apparatus

Technical field

The present invention relates generally to implantable medical device, relating to the method and apparatus for distinguishing supraventricular tachycardia (SVT) and ventricular tachycardia (VT) particularly.

Background technology

Typical Implantable Cardioverter Defibrillator (ICD) has the ability providing various anti-tachyarrhythmia pacemaker (ATP) therapy and Cardioversion/defibrillation shock therapy.Usually, depend on the ARR type detected, apply arrhythmia treatment according to from not too radical to the pre-programmed sequences of more radical treatment.Usually, by being returned to wherein coherent spontaneous R wavelength-division and leaving the rhythm of the heart of the demand pacing of at least one restriction interval or sinus rhythm interval confirming ARR termination.When ATP attempts stopping tachycardia, high pressure Cardioversion can be transmitted and impact.Because to impact patient be painful and expend relatively more battery charge compared with pacemaker impulse, therefore expect by using not too radical pacing therapy successfully to stop tachycardia to avoid transmitting the needs impacted.

The success section ground of tachycardia treatment depends on the precision that tachycardia detects.In some cases, the tachycardia (that is, supraventricular tachycardia (SVT)) coming from atrium is difficult to distinguish with the tachycardia coming from ventricle (that is, ventricular tachycardia (VT)).Such as, when SVT is transmitted to ventricle or when VT by drive in the wrong direction be transmitted to atrium time, atrium chamber and ventricle chamber both can present similar tachycardia cycle length.Therefore, need to be used for when having the highest successful probability and not having unacceptable delay to attempt when stopping tachycardia the tachycardia detected to be classified as exactly or VT or SVT event with the method allowing ICD to transmit most suitable treatment.

Summary of the invention

According to a kind of armarium for distinguishing cardiac event of the present invention, comprising: multiple electrode, sense the almost whole persistent period sterically defined heart signal across predetermined sensing window, the heart signal sensed comprises two-way amount heart signal; And processor, described processor is configured to: determine the configuration coupling mark corresponding to sensed heart signal and known template; Described configuration coupling mark is compared with multiple form Matching band; Select across the heart beating feature in multiple heart beating features of the whole persistent period being less than described sensing window, described heart beating feature based on described configuration coupling mark fall into described multiple form Matching band which and choose from multiple heart beating feature; Accumulation cardiac event evidence is carried out in response to described configuration coupling mark and determined heart beating feature; And distinguish cardiac event in response to accumulated cardiac event evidence.

Accompanying drawing explanation

Fig. 1 is the schematic diagram of implantable medical device (IMD).

Fig. 2 is the functional block diagram of the IMD according to Fig. 1 of an embodiment.

Fig. 3 is the state diagram that tachycardia detected and distinguished mode of operation included in algorithm.

Fig. 4 is the flow chart of the operation performed in the state 1 of tachycardia detection algorithm.

Fig. 5 is the flow chart of the heart rate mutating detector of low RR interval variability (variability) mode of operation in state 1.

Fig. 6 is the curve of the expection RR interval scope using RRMEAN and RRMAD of RR interval (RRI) tolerance and calculate.

Fig. 7 is the flow chart for controlling the method switched between low RRI variability pattern and high RRI variability pattern during state 1.

Fig. 8 is the flow chart of the operation performed by RRI variability change detector during high variability pattern.

Fig. 9 is the flow chart of a kind of method for comparing the heart rate estimated value obtained from two different sensing vectors.

Figure 10 is to provide the flow chart of the general view of the operation performed during the state 2 of detection algorithm.

Figure 11 is the flow chart for carrying out the method distinguished during state 2 operation between VT (can treat) and the rhythm of the heart of SVT (untreatable).

Figure 12 is the flow chart of a kind of method extracting specific heart beating characteristic sum accumulation VT evidence on the basis at heart beating one by one (beat-by-beat).

Figure 13 A is the flow chart of application SVT confidence district (confident zone) heart beating characterization rules.

Figure 13 B is the flow chart of a kind of method for applying VT heartbeat when configuration mark (morphology score) falls into SVT grey area.

Figure 13 C is the flow chart of a kind of method for applying VT grey area rule.

Figure 14 is the flow chart for the method across multiple configuration mark grey area application rule.

Figure 15 is for detecting the flow chart of the process of the rule of rhythm of the heart breakpoint (breaking point) for application.

Figure 16 is the flow chart of a kind of method for applying the rule for detecting rhythm of the heart breakpoint across whole SVT form mark district.

Figure 17 is the flow chart for adjusting the process of VT evidence count value on the basis of heart beating one by one in response to form mark district rule.

Figure 18 is the flow chart for abandoning the method in process, current heartbeat being classified as go to pot signal or the signal that do not go to pot at noise/pseudomorphism.

Figure 19 be for calculate EGM signal is classified as go to pot signal or do not go to pot signal time the flow chart of a kind of method of the tolerance of noise (such as to the relevant pseudomorphism of lead-in wire) that uses.

Figure 20 is the flow chart for the method changed between detection algorithm state.

Figure 21 is the flow chart of operator scheme after the treatment performed after reentering state 2 after transmitting tachycardia treatment.

Figure 22 is the flow chart for detecting a kind of method that the rear VT for the treatment of stops according to an embodiment.

Figure 23 is the flow chart for again detecting a kind of method of VT during operator scheme after the treatment.

Detailed description of the invention

In the following description, with reference to each illustrative embodiment.Should be appreciated that and can use other embodiments when not deviating from the scope of the present disclosure.In some instances, for clearly object, identical drawing reference numeral can be used in the accompanying drawings to identify similar element.As used herein, term " module " refers to special IC (ASIC), electronic circuit, processor (shared, special or organize processor) and the memorizer, the combinatorial circuit that perform one or more software or firmware program or provides other suitable assemblies of described function.

Disclose a kind of for detecting the tachycardia detection algorithm can treated and also carry out betwixt with the untreatable rhythm of the heart distinguishing.As used herein, term " can be treated the rhythm of the heart " and refers to and be come from ventricle and can by transmitting treatment (all anti-tachyarrhythmia pacemakers in this way or ventricular rhythm turn multiple or go impact of quivering) with any tachycardia stopping ventricular tachycardia to treat in ventricle." untreatable " rhythm of the heart is any rhythm of the heart with relatively slow Ventricular Rate (lower than ventricular tachycardia speed) and any tachycardia come from room.Only usually supraventricular tachycardia can not be alleviated in the numerous transmission treatment of ventricle chamber intermediate frequency.

As used herein, " the relevant rhythm of the heart " meets to be transformed into relevant (concerned) detected state to detect any rhythm of the heart of the criterion of potential the treated rhythm of the heart from incoherent (unconcerned) detected state.Can be different between each embodiment for being transformed into the criterion of relevant detected state, but generally include the enough fast Ventricular Rate of detection thus may to suddenly change the rhythm of the heart of (such as, the sharply increase of Ventricular Rate or the sharply reduction of RR interval (RRI) variability) for the tachycardic rhythm of the heart can be treated or there is the rhythm of the heart that can be associated with VT.

VT system refers to any fast ventricular rhythm of satisfied detection criteria as above, and does not get rid of ventricular fibrillation (VF), unless indicated clearly in addition.In illustrative embodiment described in this article, detection algorithm depends on each heart rate limit value and detects and distinguish VT and SVT." Monitoring lower-cut " cannot detect to treat tachycardic heart rate below it.Can treat the rhythm of the heart to detect, Ventricular Rate must faster than verification and measurement ratio lower limit (or RRI ratio be associated Monitoring lower-cut interval short).SVT limit value more than it, rhythm of the heart is classified as VT and is designated the heart rate can treating the rhythm of the heart.Lead the fast Ventricular Rate of limit value (or RRI is shorter than SVT limit value interval) than SVT be considered to too fast to such an extent as to can not come from room.Such as, in one embodiment, the nominal value of Monitoring lower-cut interval is about 400ms and the nominal value of SVT limit value interval is about 240ms.The heart rate characterized by the RRI grown than Monitoring lower-cut can not be the relevant rhythm of the heart.Depend on other factors, the heart rate characterized by shorter than Monitoring lower-cut but longer than SVT limit value RRI may be the relevant rhythm of the heart.The heart rate characterized by the RRI shorter than SVI interval limit value may be detected as VT.

Fig. 1 is the schematic diagram of implantable medical device (IMD) 10.IMD 10 is presented as ICD in FIG.But method described herein should not be construed as limited to any specific implantable medical device or any specific cardiac medical devices.On the contrary, each embodiment can comprise any cardiac medical devices, as long as the rhythm of the heart of patient monitored by the multiple electrode of this equipment use or other sensors.These electrodes can sense the heart EGM or ECG signal that are referred to as " heart signal " in this article.

In FIG, right atrium (RA), left atrium (LA), right ventricle (RV), left ventricle (LV) and from right atrium opening extend schematically show in heart 12 with the coronary sinus (CS) forming great cardiac vein.Article two, respectively IMD 10 is connected with RV and LV with 18 through transvenous lead 16.Each lead-in wire comprises at least one electric conductor and pace-making/sensing electrode.Such as, lead-in wire 16 and 18 is connected respectively to pace-making/sensing electrode 20,22 and 24,28.In addition, housing electrode 26 can be formed as a part for the outer surface of the shell of equipment 10.Pace-making/sensing electrode 20,22 and 24,28 and housing electrode 26 are optionally used to provide multiple one pole for pace-making and sensing function and bipolar pacing/sensing electrode combination.In left and right ventricles or shown in around left and right ventricles, position is only illustrative.In addition, other lead-in wires and pace-making/sensing electrode can be used to replace lead-in wire and electrodes shown in any one or more, or can by these other lead-in wire and pace-making/sensing electrode and these any one or more shown in go between and combination of electrodes uses.

Usually, in the pacing system of Fig. 1 shown type, the electrode being labeled as " pace-making/sensing " electrode is in this article used as pace-making and sensing two kinds of functions.In a particular embodiment, these electrodes can be specially used as programming or acquiescence combination pace-making or sensing electrode for sense heart signal and transmit pacemaker impulse.Described lead-in wire and electrode can be used for recording heart signal.Recorded data can be transmitted periodically programmable device or can to take remote measurement other external equipments communicated with IMD 10.

RV coil electrode 34 and superior vena cava (SVC) coil electrode 32 are also shown to be coupled to a part for RV lead-in wire 16.Additionally or alternatively, coil electrode can be coupled to multiple parts of CS lead-in wire 18.Coil electrode 32 and 34 or other similar electrode type can be electrically coupled to high-tension circuit to transmit high pressure Cardioversion/defibrillation shocks pulse.

Electrode shown in Fig. 1 can be arranged on each position in heart, around heart or on heart, and is not limited to shown position.In addition, other lead-in wire and systems shown in the alternative Fig. 1 of electrode system.Detection algorithm described herein does not need use to detect and distinguish can treat the rhythm of the heart for the electrode sensing atrial signal.Thus, IMD 10 is illustrated as only being coupled to ventricular lead 16 and 18, but the realization of detection algorithm is not limited to the system only adopting ventricular lead.In other embodiments, dual cavity or multichamber system can be used, these systems comprise for being placed on by electrode in the chamber of atrium, atrium lead-in wire around atrium chamber or atrium chamber.

ICD and pacemaker use the single ventricle EGM signal for sensing ventricular cardiac event (R ripple) to determine the needs of pace-making and to detect the RR interval meeting tachycardia detection criteria usually.It can be use the one pole of one or two electrode be placed on respectively in ventricular heart chamber or on ventricular heart chamber or bipolar sensing vector that EGM senses vector.The sensing error that can occur about single ventricle EGM signal can cause ICD to transmit unwanted treatment.Generable typical case's sensing error comprises crossing sensing, electromagnetic interference, noncardiac muscle potential noise, lead-in wire relevant pseudomorphism or other non-physiological noises and sensing the two of single QRS wave group (complex) T ripple.As the result of the pacing energy sent from the identical electrodes for sensing, also can comprise very unpolarized deficient sensing and the sensing error crossing sensing to T ripple.

Tachycardia detection algorithm described herein adopt to use when estimating heart rate and on the basis in heart beating one by one while application rule two-way amount EGM sensing accumulate for detecting the VT evidence can treating the rhythm of the heart.A sensing vector is selected to the EGM signal (in this article also referred to as far field (FF) signal) of EGM signal providing and have the overall situation relatively because at least one sensing one pole be placed with its space when representing that action potential signal occurs in the larger area of ventricle away from ventricle chamber to obtain and signal.Second sensing vector is selected to the EGM signal (in this article also referred to as near field (NF) signal) providing the EGM signal with relatively local because two electrodes are usually located in ventricle chamber or on ventricle chamber to obtain the ventricle EGM signal space of the action potential signal of small size (more and) of more local.

In the illustrative embodiment shown in Fig. 1, the arbitrary electrode being positioned at ventricle or Supraventricularis electrode 20,22,24,28 and 34 matched by any electrode (such as svc coil electrode 32 or housing electrode 26) used be arranged in away from ventricle place is to obtain FF signal.By selecting any two electrodes being positioned at intraventricular electrode 20,22,24,28 and 34 of bipolar centering to obtain near-field signals.Such as, NF EGM signal can be sensed between RV point electrode 22 and RV annular electrode 20.Use RV coil electrode 34 and housing electrode 26, FF EGM signal can sense with NF EGM signal simultaneously.

Expect to use two the different sensing vectors not sharing public electrode to sense FF and NF EGM signal.But, depend on used electrode and lead-line configuration, between the EGM signal that some embodiments can sense at two simultaneously, adopt public electrode.Tachycardia detection algorithm described herein relates to use FFEGM signal and NF EGM signal, but, in an alternate embodiment, tool can be used to be with or without any two different sensing vectors of common electrode, and these sensing vectors comprise at least two FF signals, at least two NF signals or comprise any combination of combination of an a FF and NF signal.

Each embodiment described herein is not limited to use intracardiac lead or transvenous lead-in wire.Electrode or the even outer electrode system of subcutaneous implantation can be used.In these cases, compared with the second pair of electrode separating relatively large distance for obtaining " far field " signal, bipolarly more closely obtain " near field " signal by being spaced.

Fig. 2 is the functional block diagram of IMD 10 according to Fig. 1 of an embodiment.IMD 10 generally comprises timing and control circuit 52 and controller, and this controller can be presented as microprocessor 54 or digital state machine for carrying out timing to sensing and treatment transmitting function according to operator scheme by programming.Microprocessor 54 and the memorizer 56 be associated are coupled to each assembly of IMD 10 via data/address bus 55.IMD10 comprise for timing and control 52 and microprocessor 54 control under by comprising cardiac pacing pulse, the electric stimulation pulse of arrhythmia pacing therapy (such as anti-tachyarrhythmia pacemaker (ATP) and Cardioversion/defibrillation shocks) is sent to the treatment delivery module 50 of patient's heart.Treatment delivery module 50 is coupled to two or more electrodes 68 via optional switch matrix 58 usually.Electrode 68 corresponds to each electrode shown in Fig. 1.Switch matrix 58 is for selecting which electrode and corresponding pole for transmitting electric stimulation pulse.

Sensing cardiac electric signals is for determining when to need electronic stimulation and controlling the sequential of boost pulse.The electrode for sensing and the electrode for stimulating can be selected via switch matrix 58.When for sensing, the heart signal that electrode 68 receives can be coupled to signal processing circuit 60 via switch matrix 58.Signal processor 60 comprises sense amplifier, and can comprise other circuit for signal conditioning of such as wave filter and analog-digital converter and so on.Then, cardiac electric signals can by microprocessor 54 for detecting the physiological event such as detecting and distinguish arrhythmia and so on.Signal processing circuit 60 comprises for sensing ventricular cardiac event (that is, R ripple) with the cardiac event sensing circuit used when determining RRI and QRS waveform morphology.

Tachycardia detection algorithm realizes with the IMD controller of the untreatable rhythm of the heart by for detecting and distinguish to treat.Sensing ventricular cardiac event interval (RRI) and R waveform state use detecting VT distinguishing in VT and SVT.Whether can be the determination can treating the rhythm of the heart without the need to using atrial signal to make about the rhythm of the heart based on ventricle EGM signal.

Can the rhythm of the heart be treated in response to detection, under timing with the control controlling 52, transmit treatment by treatment delivery module 50.Treatment can be transmitted according to treatment menu by programming.Arrhythmia treatment can comprise layering treatment menu, wherein first transmits not too radical ATP therapy, and when unsuccessful, transmit impacting with high pressure treatment.

IMD 10 can additionally be coupled to the lead-in wire extended from IMD 10 carry or be combined in one or more biosensors 70 IMD shell or on IMD shell.The signal carrying out sensor 70 is received by sensor interface 62, and sensor signal is supplied to signal processing circuit 60 by this sensor interface 62.Sensor signal can be used by the microprocessor 54 for detecting physiological event or situation.

Operating system comprises the associated memory 56 for storing the parameter value various by programming used by microprocessor 54.Memorizer 56 also can be used for storage and compiles and/or data history-sensitive with equipment operating from sensed EGM/ECG and other physiological signals, goes out for the remote measurement when receiving retrieval or inquiry instruction.Parameter and tachycardia distinguishing rule and algorithm can be stored in memorizer 56 and to be used by microprocessor 54.

IMD 10 also comprises telemetric circuit 64 and antenna 65.Program command or data are transmitted during up-link between ICD telemetric circuit 64 and the external telemetry circuit being included in programmable device or monitoring means and downlink telemetry.

Fig. 3 illustrates that tachycardia detects and distinguishes the state Figure 100 of the mode of operation that can comprise in algorithm.Tachycardia detection algorithm comprises four modes of operation 102,104,106 and 108.State 1102 is the uncorrelated states corresponding to the state that RRI monitoring wherein occurs.Perform the analysis of RRI to detect the sudden change of the rhythm of the heart.Sudden change can be the sudden change (that is, the sudden change of RRI length) of heart rate (HR) or the sudden change (that is, the sudden change of RRI difference) of RRI variability.HR change detector and RRI variability change detector operate in state 1102, and will describe in detail hereinafter.

When in state 1102 meet be applied to the abrupt climatic change criterion of the RRI measured or high heart rate criterion time, occur to the state 2104 i.e. conversion of correlation behavior.Occur from state 1 to the conversion of state 2 without the need to performing additional morphological analysis based on RRI monitoring.In order to get the hang of 2104, in state 1104, detect that HR increases, from measuring the RRI shorter than Monitoring lower-cut interval.State 2104 is " correlation behaviors ", because HR increases, but causes the heart chamber of fast Ventricular Rate may be uncertain.Need analyzing adjuncts to distinguish SVT and VT.In the process of state 2, on the basis of heart beating one by one, the morphological analysis of ventricle EGM signal is used to accumulate the evidence of VT.Except RRI analyzes, morphological analysis is used to determine that the rhythm of the heart is " can treat " the VT rhythm of the heart or " untreatable " rhythm of the heart.

Can occur to get back to state 1 (uncorrelated state) from state 2 or proceed to the conversion of state 3 (be sure oing state) based on RRI data or based on the combination of RRI data and EGM signal aspect data individually.Thus, in state 2104, RRI monitoring continues and performs the additional monitoring to EGM signal aspect, thus on the basis of heart beating one by one, accumulates the evidence of VT, as described in detail in this article.If RRI criterion and VT evidence meet VT detection criteria, then occur to the conversion of state 3106.If RRI criterion and/or VT evidence no longer meet remain on criterion required in state 2, then get back to the conversion of state 1102.

Once arrive state 3106, just detect VT and, as started therapeutic choice process according to treatment menu by programming.But, due to the beginning for the treatment of can charge because of capacitor, treatment delay by programming or other reasons and postpone, therefore IMD control system can remain in state 3 and reach an interval.The RRI monitoring performed in state 2 and morphological analysis continue in state 3.

If RRI data instruction HR drops to below related rate (that is, below verification and measurement ratio lower limit), then can occur from state 3106 directly to the conversion of state 1102.But if RRI data or morphological analysis no longer meet VT detection criteria remain on more than the threshold value of correlation behavior, then can occur to the conversion of state 2104.

When imminent treatment is ready for transmission, there is the conversion from state 3106 to state 4108.Such as, start the treatment delay of guide, capacitor charging or other times interval at the expiration as actual therapeutic, and be made to the conversion of state 4.Treatment is transmitted in state 4.After treatment transmits, get back to the conversion of state 2104 to continue to monitor the rhythm of the heart.Based on RRI criterion and morphological analysis, detect and distinguish algorithm and remain in state 2104, until draw the judgement of the state of being back to 1102 or state 3106.Now, the various State Transferring and operation that perform in each detection algorithm state will be described in greater detail in.

Fig. 4 is the flow process Figure 150 of the operation performed in the state 1 of tachycardia detection algorithm.Flow process Figure 150 and other flow charts illustrated in this article are intended to illustrate the feature operation of this equipment, and should not be interpreted as reflecting the particular form putting into practice the necessary software of described method or hardware.Can think, the particular detection that the particular form of software, firmware and/or hardware adopts primarily of the specific system architectures adopted in this equipment and this equipment and treatment transfer approach are determined.There is provided the software, firmware and/or the hardware that realize described function in the limit of power of those skilled in the art in the situation of any modern medical equipment that disclosure in this article provides.

The method described in conjunction with the flow chart presented herein can realize for making programmable processor perform in the computer-readable medium of the instruction of described method in storage." computer-readable medium " includes but not limited to any volatibility or non-volatile media, such as RAM, ROM, CD-ROM, NVRAM, EEPROM, flash memory etc.These instructions can be implemented as one or more software module, and these software modules can be performed by itself or perform with other combination of software.

After IMD initializes or after being back to state 1 from state 2 or state 3, get the hang of 1 at frame 151.At frame 152, sensing FF and NF EGM signal.Use two-way amount to sense method to allow before there is State Transferring, use the 2nd EGM signal to confirm the heart rate detected.Two-way amount sensing method also allows the overall signal form optionally analyzed in an analysis window and/or the special characteristic (being referred to herein as " heart beating feature ") of FF and NF signal analyzed in this analysis window, thus when operating in state 2 and 3 with provide the best result of SVT and VT from mode perform.

When sensing NF and FF EGM signal, pacemaker impulse is sent to heart to also can be used for by NF sensing electrode.If do not transmit ventricular pace pulse (as determined at decision box 153), then sense ensuing NF and FF event at frame 156.At frame 157, be used in current NF/FF sensed event and determine NF HR and FF HR respectively to corresponding NF and the FF RRI previously measured between NF/FF sensed event.

As from the illustrative embodiment described in detail in this article by apparent, NF EGM signal can be considered to " mainly " sensing signal because this NF signal for sense cardiac event with measure for estimate HR RRI, state 1 operate during detect the rhythm of the heart sudden change and state 2 operate during for arranging morphological analysis window." secondary " to the analysis of FF EGM signal because FF EGM signal for verify NF heart rate and after identifying the relevant rhythm of the heart for the morphological analysis in state 2.Thus, in certain embodiments, relative to NF event sensing, mark sensed event and measure about FFEGM signal sensed event between interval can postpone in time.About NF EGM signal, can cardiac event be sensed and RRI can be measured in real time.FF EGM signal can be analyzed with NF EGM signal simultaneously in real time, or buffer memory is for analyzing after a while.After determining the result needing to verify NF signal analysis, use buffered signal search FF sensed event to carry out ex post.Accurate to the real-time sensing of the ex post facto analysis possibility comparison cardiac event of the storage FFEGM signal for sensing R ripple.

If as transmitted ventricular pace pulse definitely at frame 153, then " force " FF EGM signal " sensing " cardiac event.The FF event flag used when frame 154 is automatically created on and measures RRI is to estimate the FF HR of Current heart rate at frame 156.In addition, FF is set at frame 155 and senses threshold value to prepare to sense next FF event.Usually, self-adjusting sensing threshold value by follow the tracks of sensed event amplitude with and subsequent decay.Because pacemaker impulse causes " sensing " event of the automatic generation about FF signal, therefore use the replacement method for arranging self-adjusting sensing threshold value when transmitting pacemaker impulse.

When transmitting pacemaker impulse, be blank phase (blankingperiod) after the FF event flag automatically generated.Then, FF senses the amplitude that threshold value follows the tracks of induced response within the whole FF EGM blank phase, but does not sense new events during the blank phase.When the blank phase expires, sense at frame 155, FF the percentage ratio that threshold value is set to the peak amplitude followed the tracks of during the blank phase.Subsequently, FF senses threshold value and decays, until transmit another pacemaker impulse or until sense the intrinsic event about FF EGM signal along with the time.

At frame 157, be used in each NF and FF RRI of measuring between sensed event to measure NF HR and FF HR.If the NF HR as determined at frame 158 is greater than sudden change limit value and NF and FF HR roughly mates (frame 168), then occur to the conversion of state 2 at frame 170.At the corresponding RRI that the sudden change limit value of frame 158 use is threshold value heart rate or drops between verification and measurement ratio lower limit and SVT limit value.The heart rate (or than sudden change limit value interval short RRI) larger than sudden change limit value is identified as " the relevant rhythm of the heart ", and without the need to detecting the sudden change of the rhythm of the heart.Particularly, detection algorithm does not need the sudden change observing HR or RRI variability to identify the relevant rhythm of the heart.Such as, the limit value that suddenlys change can be the HR of 190 jumpings/(bpm) per minute.At more than 190bpm, suddenly change to state 2 from state 1 and do not need abrupt climatic change requirement, because very high HR itself is considered to the relevant rhythm of the heart.

In order to current HR and HR threshold value (such as suddenly change limit value or Monitoring lower-cut, SVT limit value or other heart rate thresholds described herein) be compared, multiple method can be used.One method comprises the intermediate value of determining recently " m " individual RRI and this intermediate value and threshold value heart rate is compared.Other method requires that specific quantity " n " the individual RRI in nearest " m " individual RRI is shorter than the interval corresponding with threshold value HR.

In one embodiment, collect predetermined quantity " m " individual continuous RRI, and " m " individual between interim the n-th minimum RRI be used as to compare for HR threshold value the estimation of current HR.Such as, the 9th minimum RRI in nearest 12 RRI can be used as comparing for HR threshold value the estimation of current HR.At frame 157, if the 9th minimum RRI is shorter than sudden change limit value interval, then detection algorithm advances to decision box 168.

Use the n-th minimum RRI in the collection RRI of predetermined quantity can produce the result different from using RRI intermediate value as the method for the tolerance of HR.When being in use worth, deficient sensed event destroys a RRI (by creating a very long RRI).Cross sensed event and destroy two RRI (by creating two very short RRI).Therefore, compared with deficient sensing, crossing sensing can cause destroyed intermediate value quickly.In practice, in modern IC D, cross sensing usually more usually to occur than deficient sensing.Cross sensing and can cause over-evaluating HR, and can cause unnecessarily being transformed into state 2.With using RRI intermediate value compared with the method for the tolerance of HR, by select specified quantity nearest continuous RRI in the n-th the shortest RRI, by excessivelys sense the state that is transformed into 2 caused probability minimizing.

When NF HR estimated value exceedes mutation rate limit value, the NF HR of frame 168 and the comparison of FF HR can be comprised to confirm the NF HR detected.The comparison of frame 168 can comprise the determination relevant to the reliability of FF EGM signal to guarantee that FF signal is not noise or destroyed pseudomorphism and signal intensity is reliable.If the sensed event of predetermined percentage at minimum cardiac event sensing threshold value place or in its vicinity or have peak amplitude lower than predetermined reliable thresholds amplitude place higher than event sensing threshold value, then can be considered to insecure for the arbitrary signal estimated for heart rate in NF or FF signal.Have and almost can not reach sensing threshold value or be sensing or the concern that other events sensed can occur to owe than being arranged to can promote these sensed event a little more than the altofrequency sensed event of the low amplitude of reliable amplitude threshold of event sensing threshold value.If not yet there is sensed event within deficient sensing threshold value interval, then EGM signal can be additionally or alternatively defined as unreliable.Such as, if having passed through at least 2 seconds or the longer time between two sensed event, then EGM signal can be classified as estimation HR has been insecure.

When not by crossing sensing, owe appreciable error that sensing, noise artifact or other factors cause, NF and FF RRI measurements of combination can be used in every way to verify the heart rate accurately sensed.In one embodiment, also to exist within time predefined interval from NF sensed event (such as by checking for each sensed event about NF EGM signal, in about 20ms) the corresponding sense event about FF EGM signal that occurs, FF HR can be used to verify NF HR at frame 168.

As be applied to NF class signal seemingly, the method for the tolerance determining HR can be applicable to FF signal, the above-mentioned example of the n-th minimum RRI in such as nearest multiple continuous RRI.If FF HR estimated value is not roughly mated with NF HR estimated value, e.g., in matching range, then detection algorithm remains in state 1.Specifically, if FF HR estimated value is less than sudden change limit value, then rely on the lower HR estimated value based on FF HR but not NF HR estimated value.This process returns frame 152 to sense ensuing FF and NF event and to measure next RRI about FF and NF EGM signal.Although do not illustrate clearly in the diagram, if FF HR estimated value is different significantly but than NF HR fast instead of slower than it (such as, fast than SVT limit value), then can be made to the conversion of state 2 at frame 170.

If the HR estimated value from FF EGM is different from the heart rate estimated value of NF EGM significantly, then at frame 169, making during this is determined buffer memory RRI and/or FF that use and NF EGM signal segment can be stored as diagnostic data to be recorded in not mating of occurring between FF and NF HR.This diagnosis may contribute to clinician and technical staff determines when and why two-way amount EGM signal data is contradiction and takes corrective action as required in HR estimated value, such as contradictory outcome show as be because noise, owe the relevant situation of sensing, lead-in wire or other non-physiological reasons time.

As long as HR keeps below sudden change limit value (be negative decision at frame 158), HR mutating detector 160 or RRI variability mutating detector 164 will operate in state 1.Operate in low variability (LV) pattern or high variability (HV) pattern, as long as NF HR keeps below mutation rate limit value at frame 159, IMD controller determination detection algorithm.Composition graphs 7 is described frame 159 to the control switched between LV pattern and HV pattern.

In LV pattern, HR change detector operates the sudden change to detect HR based on the abrupt climatic change criterion applied at decision box 162 at frame 160.Hereinafter, the details relevant to the operation of the HR mutating detector of frame 160 is described composition graphs 5.

If meet HR at frame 162 to suddenly change criterion and NF HR is greater than Monitoring lower-cut (frame 167) and if NF and FF HR roughly mates (frame 168), then occur to the conversion of state 2 at frame 170.The Monitoring lower-cut applied at frame 167 is the limit value being applied to HR, even if detect that below this limit value the sudden change rhythm of the heart of HR or RRI variability neither be correlated with the rhythm of the heart.Such as, Monitoring lower-cut nominally can be set to 150bpm.If be less than Monitoring lower-cut at frame 167HR, then detection algorithm to remain in state 1 and is back to frame 156 to proceed to next sensed event.Until the HR estimated value about current heartbeat obtained at least meets Monitoring lower-cut requirement, detection algorithm just proceeds to state 2.

If detection algorithm operates in HV pattern (be negative decision at frame 159), then RRI variability change detector operates at frame 164.If HR is that heart beating alterable height is (such as one by one, during the rhythm of the heart of atrial fibrillation (AF), frequent dystopy, unstable intrinsic activation, premature ventricular beat (bigemeny), premature ventricular beat (trigemeny) or other alterable heights), then the sudden change of HR that labelling VT starts can be covered by the RRI of alterable height.

Thus, when RRI be one by one heart beating alterable height time, tachycardia is distinguished algorithm during state 1, is operated the detection of the sudden change realized RRI variability in HV pattern.Generally speaking, if the rhythm of the heart that VT is characterized by the RRI of alterable height causes, then the unexpected reduction of RRI variability will be there is.When the criterion (frame 166), the NF HR that meet the sudden change for detecting RRI variability are greater than Monitoring lower-cut (frame 167), and when NF and FF HR roughly mates (frame 168), occur to the conversion of state 2 at frame 170.Hereinafter, composition graphs 8 is described the details relevant to the operation of the RRI variability mutating detector at frame 164.

Fig. 5 is flow process Figure 200 of the HR mutating detector of LV mode of operation in state 1.HR mutating detector can operation in replacement pattern 202 or normal mode 210.After device initialize or in response to the manual replacement of this algorithm, replacement pattern 202 operates.During replacement pattern 202, at frame 204, initialize two tolerance of the RRI from NF EGM signal.Average RRI (RRMEAN) is initialized to nominal value, such as 900ms.In addition, the expection absolute difference between next RRI and RRMEAN is initialized.This expection absolute difference (being called RRMAD) defines expection or predicts the scope around the RRMEAD that next RRI falls into.The nominal initial value of RRMAD is about 800ms.

Based on the RRI of actual measurement, use down " n " individual RRI so that RRMEAN and RRMAD is updated to actual value.Based on each new RRI, RRMEAN can be calculated as the weighted sum of last RRMEAN and current RRI.Such as, RRMEAN (through upgrading)=0.5 (RRMEAN)+0.5 (RRI _current).Although employ equal weight coefficient in above-mentioned equation, will be appreciated that other weight coefficients during being used in replacement pattern 202 rapidly close to expection RRMEAN value.

Each new RRI is used RRMAD to be updated to the weighted sum of the currency of RRMAD and the difference between the currency of RRI and RRMEAN of pre-test.Such as, RRMAD can upgrade according to following equation:

RRMAD (through upgrading)=0.5 (RRMAD)+0.5 (│ RRI _current– RRMEAN │+k*RRMEAN)

The minimal size of " k*RRMEAN " item to RRMAD imposes restriction, and wherein " k " is minimum fixed value or percentage ratio (such as, being less than about 0.05).Alternatively, RRMAD can retrain by fixing minima.

Quantity " n " for adjusting the RRI of RRMEAN and RRMAD during replacement pattern 202 in the scope of about 3 to 8, but can be not limited thereto.In one embodiment, at frame 206, use the first five RRI to calculate RRMEAN and RRMAD for defining expection RRI scope.If current RRI does not fall into RRMEAN ± RRMAD, be then not used at replacement pattern 202 period RRI and upgrade RRMEAN or RRMAD and be not counted as one of " n " individual RRI adjusting RRMEAN and RRMAD." n " individual RRI is by RRMEAN and RRMAD from after initial nominal values is adapted to actual desired value rapidly before use, and this algorithm enters normal operating 210.During normal operating 210, use RRMEAN and RRMAD value to calculate expection RRI scope at frame 212:

RRI (expection)=RRMEAN ± RRMAD

When sensing next NF event (frame 214), measuring RRI at frame 216 and itself and the RRI scope of expecting are compared.If RRI is in desired extent, then at frame 217, this RRI is for upgrading RRMEAN and RRMAD.The formula used during resetting operator scheme 202 can be different from for the formula upgrading RRMEAN and RRMAD at frame 217.Such as, different weight coefficients can be used and/or can addition Item be comprised.In one embodiment, use the relatively low weight coefficient being applied to current RRI to calculate RRMEAN, such as:

RRMEAN (through upgrading)=0.9 (RRMEAN)+0.1 (RRI _current).

RRMAD can be calculated as:

RRMAD (through upgrading)=0.95 (RRMAD)+0.05* (d) * (│ RRI _current– RRMEAN │+k*RRMEAN) * { 1 – ((750 – RRMEAN)/1000) }

Wherein constant " d " is for the factor selected by stable desired extent, comprise the item " k*RRMEAN " of the minimal size for retraining RRMAD, and the factor { 1 – ((750 – RRMEAN)/1000) } forces desired extent increase along with HR and tighten up and reduce along with HR and expand.In other words, expect that RRI scope narrows at higher HR place and broadens at lower HR place.The largest amount of definable RRMAD retrains greatest expected scope.In one embodiment, RRMAD is limited to the largest percentage of RRMEAN, about 20% of such as RRMEAN.

If RRI is beyond desired extent (RRMEAN ± RRMAD), then this RRI is not used in the updated value calculating RRMEAN and RRMAD.At frame 220, upgrade and lose enumerator thus the quantity being not used in the RRI upgrading desired extent tolerance is counted.If current RRI is greater than RRMEAN+RRMAD, then loses count value and add one.If current RRI is less than RRMEAN – RRMAD, then loses count value and subtract one.

Lose enumerator can have miss-threshold on the occasion of and negative value between the value that changes.When RRI to fall in desired extent (frame 216) and for upgrading RRMEAN and RRMAD (frame 217) time, at frame 218, upgrade this loss enumerator by the value of losing enumerator is shifted near a step from its any currency to zero.

When losing count value and increasing in response to the outer RRI of scope or reduce (frame 220), at frame 222, this loss count value and miss-threshold are compared.If reached miss-threshold before changing to state 2, then state that expection RRI scope is lost at frame 224.At frame 226, by the value of RRMEAN being adjusted up or down a percentage ratio of RRMAD, upgrade expection RRI scope according to the value of losing enumerator.

In one embodiment, if current RRI is greater than desired extent and reach positive miss-threshold, then increase 25% of RRMAD to upgrade expection RRI scope at frame 226, RRMEAN.If current RRI is less than desired extent and reach negative miss-threshold, then reduce 25% of RRMAD at frame 226, RRMEAN.In this way, the skew of RRMEAN (or as required, the multiple skews in RRMEAN) reconfigures (reposition) and expects that RRI scope is with the RRI scope making the scope reconfigured again represent current expection.When HR keeps being less than sudden change limit value and not meeting other abrupt climatic change criterions for being transformed into state 2, can carry out expecting that this of RRI scope reconfigures.This algorithm to remain in state 1 and is back to frame 214 to sense next NF event.

If not yet meet miss-threshold at frame 222, then at the frame 228 outer enumerator of new range.If RRI is short (that is, being less than RRMEAN – RRMAD) unexpectedly, then the outer count value of scope adds one.Otherwise if RRI is in desired extent or be greater than desired extent, then the outer count value of scope subtracts two.The outer enumerator of scope is used for counting the quantity of RRI shorter than expection RRI scope all the time.If this counting is tending towards upwards, then there is the RRI shorter all the time of the probability representing HR sudden change.The outer count value of scope has minimum limit value zero-sum threshold limit value, such as 20.

At frame 230, outer for scope count value and threshold count are compared the sudden change detecting HR.If not yet reach abrupt climatic change threshold value and this algorithm still operates (as determined at frame 232) in LV pattern, then this algorithm to remain in state 1 and this process is back to frame 212.In the case, expection RRI scope will keep identical, because current RRI is outer and be not used in new RRMEAN and the RRMAD value of calculating in scope.

If exceed abrupt climatic change threshold value (such as threshold value is 10) in the outer count value of frame 230 scope, represent that recent RRI is shorter than desired extent all the time, then at frame 234, NF HR and Monitoring lower-cut are compared.As described earlier, NF HR can be estimated as the n-th minimum RRI in the nearest RRI of specified quantity.If NF HR estimated value does not exceed for detecting tachycardic verification and measurement ratio lower limit and this algorithm is still in (frame 232) in LV pattern, then this process is back to frame 212.Expection RRI scope will keep identical, again because current RRI is outside scope.Even if detected that HR suddenlys change, the too low to such an extent as to relevant rhythm of the heart that can not be considered to for tachycardia testing goal of this HR.

If as NF HR is faster than Monitoring lower-cut definitely at frame 234, then check that HR from FF EGM signal is to verify NF HR.At frame 236, first analyze FF EGM signal to determine that whether signal is reliable.Because FF EGM signal is more easily subject to the impact of noise or pseudomorphism, therefore noise/pseudomorphism abandons criterion and can be applicable to FF EGM signal and be defined as insecure FF HR data to abandon.In order to be considered to " reliably ", FF EGM also must have enough amplitudes.In other words, if FF R ripple signal has very little amplitude, then the HR estimated value from FF EGM is considered to insecure and is not used to checking or counterevidence NFHR.

At frame 236, the specified criteria relevant to EGM signal amplitude and/or FF RRI can be applicable to FF signal to verify that it is the reliable signal for estimating HR.Such as, if the R ripple of the unacceptable amount in the nearest R ripple of predetermined quantity (about FF EGM signal sensing) is less than threshold amplitude and/or there is very short or very long RRI, then FF EGM signal can be confirmed as unreliable.In a particular example, if the 3rd RRI the shortest in nearest 12 RRI is less than 500ms and nearest 12 at least 4 sensing R ripples sensed in R ripples have the peak amplitude (at least one the R ripple wherein in these short arcs R ripple is in nearest three sensing R ripples) being less than 500 millivolts, then FF EGM is confirmed as insecure signal.In addition, if not yet there is sensed event (instruction low-amplitude signal and possible deficient sensing) in a certain maximum time limit value, then FF EGM signal can be confirmed as unreliable.Such as, if having passed through at least about 2500ms from nearest FF sensed event, then FF EGM signal be confirmed as unreliable.

If it is unreliable that FF EGM signal is confirmed as, then increase the unreliable count value of FF at frame 238.Then, at frame 244, rely on the NF EGM signal evidence of the relevant rhythm of the heart to be transformed into state 2, state of namely " being correlated with ".When detecting the relevant rhythm of the heart, FF EGM is not used in checking or counterevidence NF EGM result.

In one embodiment, if meet unreliable signal criterion at frame 236, then FF signal keeps classifying as unreliablely reaching a time period.For this reason, when meeting unreliable criterion, be set to maximum in the unreliable count value of frame 238, FF, such as 12.The value identical with NF HR estimated value is set in frame 239, FF HR estimated value.Thus, certainly will mate at decision box 243, NF and FF heart rate, and at frame 244, occur to the conversion of state 2.

When being found to be reliable at frame 236FF EGM signal, subtract one in the unreliable count value of frame 240, FF.As long as frame 242 definitely the unreliable count value of FF keep being greater than zero, FF EGM signal and being just considered to insecure.Be set to equal NF HR estimated value in frame 239, FF HR estimated value, thus cause the Auto-matching at NF and the FF HR of frame 234 and the conversion to state 2 at frame 244.

Although do not illustrate clearly in provided flow chart, can perform the similar analysis of NF EGM signal to determine when NF signal is insecure.If NF EGM signal is found to be insecure, then FF EGM signal can be used as sensing cardiac event and arranging the main sensing signal of morphological analysis window, until NF EGM signal is again found to be reliable.

If reach zero in the unreliable count value of frame 242FF, then estimate FF HR at frame 243 and itself and NF HR are compared.Can use and be similar to for estimating that the method as above of NF HR (that is, the n-th RRI in " m " individual nearest RRI) is to estimate FF HR.When FF HR estimated value roughly mates (frame 243) with NF HR estimated value or when FF HR estimated value exceed suddenly change limit value time, confirm determination to the relevant rhythm of the heart by FF signal.The conversion of state 2 is occurred at frame 244.HR coupling can be defined as NF HR estimated value and FF HR estimated value within the scope of predefine each other or in percentage ratio.

On the other hand, if frame 243FF EGM signal be reliably, FF HR estimated value roughly not mate with NF HR estimated value and to be not more than sudden change limit value, then do not occur to the conversion of state 2.This algorithm is back to frame 232.

At frame 232, if there is the switching from LV pattern to HV pattern, then at frame 242, be switched to RRI variability mutating detector for the method detecting rhythm of the heart sudden change from HR mutating detector.Otherwise detection algorithm to remain in LV pattern and is back to frame 212.For current heartbeat, expection RRI scope will keep identical, and at frame 214, this process will proceed to next NF sensed event.

Fig. 6 is the curve chart using RRMEAN and RRMAD of RRI to measure the expection RRI scope calculated as mentioned above.The RRI measured represents by empty circles, and draws in seconds along x-axis in units of ms along y-axis along with the time.RRMEAN 182 calculates according to the RRI measured, and illustrates with solid line.Positive boundary 184a and negative edge 184b (being referred to as 184) is limited by ± RRMAD and is illustrated by the dotted line above and below RRMEAN 182.Border 184a and 184b limits and expected based on former RRI the RRI scope that next RRI falls into.

Distribute the initial value 186 of 900ms to RRMEAN, and distribute the initial value of 800ms to RRMAD.During replacement operator scheme 188, by using the above equation provided, the first five RRI is used for being focused at actual RRMEAN value and actual RRMAD value rapidly.The normal operating 190a of the state 1 of detection algorithm starts after replacement pattern 188.At first, there is the RRI 191 of a string consistent 1000ms.RRMEAN 182 follows the tracks of consistent RRI.± RRMAD coboundary and lower boundary 184 little by little tighten up expection RRI scope is narrowed around RRMEAN.

The interval of variable RRI 192 by widening ± RRMAD border 184 makes expection RRI range expansion, as can be seen.After variable RRI 192, there is a series of two-forty RRI 193.But it is longer than the nominal Monitoring lower-cut interval of 500ms that these RRI shorten suddenly maintenance, thus State Transferring does not occur.When generation rate suddenlys change, shorter RRI is beyond expection RRI scope.Do not adjust in response to extraneous RRI, RRMEAN and RRMAD, and present all the time in the flat response of extraneous RRI.

All the time after extraneous ten RRI, lose count value and reach threshold count to reconfigure expection RRI scope.Missing mode 194 operates to reconfigure expection RRI scope.In this example, RRI is less than desired extent all the time, but HR is still lower than verification and measurement ratio lower limit, and therefore this algorithm remains in state 2.When current RRI is less than expection RRI scope, reconfigure expection RRI scope, until current RRI falls into desired extent by RRMEAN being adjusted a predetermined decrement on the basis of heart beating one by one.

If RRI is greater than expection RRI scope all the time, RRMEAN can increase predetermined increment to reconfigure expection RRI scope about each RRI.In one embodiment, when reach lose count threshold time, RRMEAN reduces as required or increases the percentage ratio (such as, RRMAD 25%) of currency of RRMAD.Expection RRI scope is allowed to reconfigure rapidly to comprise current RRI to this adjustment of RRMEAN.After this, enabling 190b.

When each RRI falls into desired extent during normal operating 190b, upgrade RRMEAN and RRMAD according to normal operating equation.Increase to 900ms at 195, RRI, it is still within the scope of the expection RRI such as limited by+RRMAD border 184.RRMEAN and RRMAD continues on the basis of heart beating one by one, use equation described above to adjust.Observe tightening up gradually of expection RRI scope.

Fig. 7 is the flow chart 250 of the method for controlling the switching between LV and HV operator scheme during state 1.As described earlier, during LV pattern, HR mutating detector operates.During HV pattern, RRI variability mutating detector operates.In order to control to operate in any pattern of detection algorithm in LV and HV pattern, the tolerance of monitoring RRI variability, i.e. " MEANVAR ".

In flow chart 250, replacement operation 202 and normal operating 210 correspond to the replacement operation shown in Fig. 5 and normal operating.Thus, the process shown in the Fig. 7 occurred during replacement operation 202 and during normal operating 210 is additional to composition graphs 5 in the process reset and describe during normal operating.

During replacement operation 202, except calculating the initial value of RRMEAN and RRMAD, for measure about " n " individual RRI before NF EGM signal, calculate the initial value of MEANVAR.MEANVAR is calculated as the weighted sum of the currency of MEANVAR and the difference between current RRI and last RRI.Such as:

MEANVAR (through upgrading)=W ₁(MEANVAR)+W ₂(│ RRI _i– RRI _i-1│)

During replacement pattern, initial nominal values can be distributed to MEANVAR, use above equation to be updated to actual MEANVAR subsequently.During resetting, W ₁and W ₂the equal value of 0.5 can be configured to.

During normal operating 210, gather NF EGM signal to sense next NF event at frame 260.In the normal operation period, the weighted sum of current MEANVAR value and the difference between current RRI and last RRI can be used to calculate MEANVAR.In one embodiment, current MEANVAR can be multiplied by the weight coefficient W1 of about 0.9 or higher, and current RRI difference can be multiplied by the weight coefficient W2 of about 0.1 or lower.Such as, W1 is about 0.96 and W2 is about 0.04.

Maximum upper limit can be applicable to MEANVAR, and it can be defined as a percentage ratio of RRMEAN, about 25% of such as RRMEAN.In order to limit the impact of outlier (outlier), limit value also can be applicable to the current RRI difference for calculating MEANVAR.Such as, if current RRI difference is greater than the currency of MEANVAR, then current RRI difference is substituted by the greater sum in MEANVAR and MEANVAR and 20ms.

At frame 262, in response to NF event, upgrade MEANVAR.If as frame 264 definitely MEANVAR be greater than a HV threshold value, then at frame 266, detection algorithm is switched to HV pattern.Be used for detecting the operation of sudden change during composition graphs 8 is described in HV pattern.

After entering HV pattern, at frame 267 Calculation Basis HR estimated value and OLD RRI AVG (old RRI AVG).OLD RRI AVG can be calculated as the meansigma methods of nearest RRI (such as, 12 to 16 nearest RRI), and is the measurement to heart rate after entering HV pattern.As will be hereinafter described in greater detail, when there is RRI variability and reducing, the trend that OLD RRI AVG increases for detecting HR.Be attended by the relevant rhythm of the heart of this combination instruction of the unexpected reduction of the RRI variability of the trend that HR increases, and can the State Transferring of detection trigger algorithm, as described below.The trend increased along with the HR of the sudden change of RRI variability not necessarily demand fulfillment as the criterion increased for " suddenly " that detect HR needed for LV pattern period HR mutating detector.In HV pattern, HR increases the criterion that can meet " sudden change " for detecting the rhythm of the heart in conjunction with the unexpected reduction of RRI variability gradually.

During HV pattern, sense next NF event at frame 268, and at frame 270, this NF event is used for upgrading MEANVAR.If as frame 272 definitely MEANVAR drop to below the 2nd LV threshold value, then at frame 274, detection algorithm is switched to LV pattern.A percentage ratio or the fixed value of RRMEAN is defined as the HV threshold value (THRESHOLD1) being switched to HV pattern from LV pattern at decision box 264.In one embodiment, THRESHOLD1 is provided in the scope of about 10 five to two ten of RRMEAN.In addition, the fixing maximum upper limit of the MEANVAR during definable LV pattern, occurs to the switching of HV pattern more than this upper limit.Such as, if MEANVAR is greater than about 0.18*RRMEAN or is greater than the fixed upper limit of 100ms, then detection algorithm is switched to HV pattern.

The 2nd LV threshold value being used for switching back LV pattern at decision box 272 can be defined as with first threshold identical or different.In one embodiment, being applied to MEANVAR is configured to lower than the first threshold for being switched to HV pattern with the THRESHOLD2 switching back LV pattern.Such as, the threshold value for switching back LV pattern can be defined as between about 10 to percent ten five of RRMEAN.In addition, the fixing lower limit of the MEANVAR during definable HV pattern, occurs to the switching of LV pattern below this lower limit.In one embodiment, if MEANVAR is less than 0.12*RRMEAN or is less than the lower limit of about 70ms, then detection algorithm is switched to LV pattern from HV pattern.The incision of HV pattern and delayed between cutting out reduce switching frequency.The upper and lower bound being applied to MEANVAR allows to switch in when variability becomes very low or very high independent of the currency of RRMEAN and occurs.

After frame 274 is switched to LV pattern, the half of LV count value and abrupt climatic change threshold value (being also referred to as " sudden change threshold value ") is compared.As described by composition graphs 8, LV enumerator is used for counting the quantity of the RRI during HV operator scheme with lower heart beating variability one by one.When exceeding sudden change threshold value in HV pattern period LV count value, the sudden change of RRI variability can be detected.Be applied to LV counting (during HV) sudden change threshold value and be applied to RRI scope count the sudden change threshold value of (during LV) outward can be identical, or different values can be arranged to during different HV and LV operator schemes.

The MEANVAR tolerance of RRI variability differently uses with LV enumerator.MEANVAR tolerance is used for the switching in state 1 between control HV pattern and LV pattern, and LV enumerator is for detecting the unexpected reduction of RRI variability, and causes being transformed into state 2 from state 1 during HV pattern.

Cause being switched to LV pattern from HV pattern at frame 274 if MEANVAR has been reduced to, then at frame 276, the currency of LV enumerator after the switching of HV pattern and a percentage ratio (such as, the half of abrupt climatic change threshold value) of abrupt climatic change threshold value are compared.If the currency of LV enumerator is not more than the selected percentage ratio of abrupt climatic change threshold value, then LV operator scheme continues, as described in conjunction with Figure 5 hereinbefore.But, if LV count value higher (such as, reaching at least half of abrupt climatic change threshold value), then calculate NEW RRI AVG (new RRI AVG) to determine whether this reduction of RRI variability is also attended by the increase of HR at frame 278.NEW RRI AVG can be calculated as the meansigma methods of the nearest RRI of predetermined quantity.Such as, recur until and nearest eight RRI (or another quantity) comprising the RRI causing being switched to LV pattern can be used for calculating NEW RRI AVG.

At frame 280, by this NEW RRI AVG and benchmark HR estimated value (namely, at the OLD RRI AVG that frame 267 calculates after entering HV pattern) compare, thus determine whether to cut out the conversion of LV pattern also along with the trend that HR increases from HV pattern.At frame 280, the ratio between NEW RRI AVG and OLD RRI AVG or difference and the threshold value being used for detecting the evidence increasing HR can be compared.Such as, if NEW RRI AVG is less than OLD RRI AVG 80 about percent (or another percentage ratio), then the HR of increase is detected.

If occur to the change of LV pattern when LV counting is greater than the half of abrupt climatic change threshold value but be not attended by the HR (frame 280) of increase, then detection algorithm continues normally to operate (being back to frame 260) in LV pattern.But if detect that HR increases, then this increases the sudden change that can indicate the rhythm of the heart together with the conversion to LV pattern when higher LV counts, thus ensures the sudden change of correlation behavior 2.

At frame 284, in order to meet State Transferring criterion, the outer enumerator of scope is set to the above threshold value for detecting sudden change.Such as, if be set as 10 for the threshold count value detecting sudden change, then at frame 284, the outer count value of scope can be set as 16.This upper count values meets the State Transferring requirement being applied to the outer enumerator of scope during LV pattern immediately.Outer for scope enumerator is set to high level allows the time analyzing FF EGM signal to verify the coupling between FF and NF HR estimated value, and this can be the requirement added before being transformed into state 2.

At frame 288, calculate FF HR estimated value and itself and NF HR estimated value are compared.Generally may correspond to the frame 236 to 243 in Fig. 5 in the process of frame 288 execution, wherein first determine the reliability of FF EGM signal, and if it is reliable for finding, then determine FF HR estimated value.If NF and FF HR estimated value matches or FF HR estimated value is greater than sudden change limit value, then occur to the conversion of state 2 at frame 290.If FF EGM signal is insecure, then depend on the result of NF signal analysis at frame 290 to affect State Transferring.Otherwise, but when FF signal be found to be reliably FF HR do not mate with NF heart rate estimated value time, do not occur to the conversion of state 2.This process is back to frame 210 and remains in the LV pattern of state 1.

Fig. 8 is the flow chart 300 of the operation performed by RRI variability mutating detector during HV pattern.At frame 302 from after LV pattern is switched to HV pattern, at frame 304, LV enumerator is reset to zero by detection algorithm.LV enumerator counts for contrasting little RRI difference desired during HV operator scheme.The RRI variability that the higher counting indicator less than expection RRI difference is lower all the time.The change of the rhythm of the heart changed from the RRI of alterable height to the less RRI of heart beating one by one can be associated with VT, and therefore can be identified as the relevant rhythm of the heart, particularly when being attended by HR and increasing.

After entering HV pattern, calculate the tolerance (frame 305) of average HR, i.e. OLD RRI AVG.Whether this tolerance OLD RRI AVG is used as benchmark HR and reduces along with RRI variability to determine that HR increases.During HV operator scheme in state 1, the combination from high RRI variability to the low sudden change of RRI variability and the HR of increase is used to the sudden change detecting the rhythm of the heart.OLD RRI AVG can be calculated as the average of nearest RRI (such as, nearest 16 continuous RRI).

At frame 306, sense next NF event, and use it for renewal MEANVAR at frame 308.As described in connection with figure 7, MEANVAR is for following the tracks of RRI variability to control the switching between HV and LV pattern, and MEANVAR initializes during the replacement operation 202 shown in Fig. 7.MEANVAR also uses in conjunction with LV enumerator with the sudden change detecting RRI variability during HV pattern.

At frame 310, the absolute difference (that is, current RRI difference) between current RRI and last RRI and low variability threshold value are compared.The low variability threshold value being applied to RRI difference can be defined as a percentage ratio of MEANVAR, thus current RRI variability and expection variability are measured and compared.In one embodiment, the low variability threshold value being applied to the RRI difference of heart beating is one by one 0.5*MEANVAR.Low variability threshold value can comprise the absolute limits of heart beating variability one by one.Such as, if current RRI difference is less than 0.5*MEANVAR or is less than about 30ms, then current RRI difference can be considered to low.

If this difference is less than low variability threshold value (evidence of low RRI variability), then at frame 318, the currency of MEANVAR and minimum MEANVAR threshold value are compared.If current MEANVAR is greater than minimum threshold, then increase in frame 320, LV count value.Comprise the requirement that MEANVAR is greater than minimum threshold, to make just to occur when average variability has been greater than certain minimum level the detection of the remarkable reduction of RRI variability, to start this detection.If MEANVAR is higher than minimum level, then less than LV threshold value current RRI difference represents the potential unexpected reduction of RRI variability.When MEANVAR is greater than minimum threshold (frame 318), increases in frame 320LV count value and be less than LV threshold value (frame 310) to make the counting of RRI difference be maintained.The RRI difference meeting these criterions provides the evidence from high RRI variability to the sudden change of low RRI variability.

If current RRI difference lower (that is, less than LV threshold value at frame 310) and MEANVAR is also less than minimum threshold (such as, being less than about 20ms), then can reduce in frame 319, LV count value.The MEANVAR being calculated as the moving average (running mean) of RRI variability is not yet high enough to the reduction detecting RRI variability.Detection algorithm remains in the HV pattern in state 1, and is back to frame 306 to sense next NF event.

When increasing in frame 320LV count value, at frame 322, itself and sudden change threshold value are compared.If LV counting is greater than sudden change threshold value, then calculate the tolerance of NEW RRI AVG as current HR at frame 323.NEW RRI AVG can be calculated as the average of nearest RRI (such as, recurring until and comprise nearest eight RRI making LV count value exceed the RRI of sudden change threshold value).

NEW RRI AVG and benchmark OLD RRI AVG is compared to determine the trend whether sudden change of RRI variability is attended by HR and increases.In one embodiment, the ratio of NEW RRI AVG and OLD RRIAVG or difference therebetween and threshold value are compared.Such as, if NEW RRI AVG is less than 90 about percent of OLD RRI AVG, then detect the evidence of the HR increased at frame 324.Will be appreciated that, estimate the additive method of current HR and benchmark HR and replace the evidence that threshold value criterion can be used for detecting the HR of increase be associated with the sudden change of RRI variability.

If the evidence of the HR of the increase of the sudden change along with RRI variability detected at frame 324, then detect the sudden change of the rhythm of the heart at frame 325.There is never correlation behavior 1 at frame 326 and arrive the conversion of correlation behavior 2.If LV enumerator not yet reaches sudden change threshold value (frame 322) or HR not yet increases to the criterion (frame 324) meeting and increase for detecting HR, then this algorithm is back to frame 306 to sense next NF event.

Referring again to frame 310, if the difference between current RRI and last RRI is greater than low variability threshold value (evidence of the high RRI variability kept), then at frame 312, this RRI difference and maximum variability threshold value are compared.If RRI difference is greater than maximum variability threshold value (such as, being greater than about 90ms), being then reset at frame 314, LV enumerator is zero.Detection algorithm remains in the HV operator scheme in state 1, and is back to frame 305 to calculate the updated value of OLD RRI AVG.

In certain embodiments, whenever LV enumerator be reset be zero time, upgrade OLD RRI AVG.In other embodiments, at frame 305, the additional criteria relevant to the behavior of RRI difference may be needed before renewal OLD RRI AVG value.Such as, before renewal OLD RRI AVG value, the continuous RRI difference of the predetermined quantity larger than maximum variability threshold value may be needed.Another criterion that may need before upgrading OLD RRI AVG is that MEANVAR is greater than MEANVAR minima.

In one embodiment, when LV enumerator be reset be zero and at least four continuous RRI differences have exceeded maximum variability threshold value (such as, 90ms) time, calculate OLD RRI AVG.Alternatively, when continuous print RRI difference at least meet another comparatively Low threshold (such as, a percentage ratio of 30ms or MEANVAR) and MEANVAR is at least greater than MEANVAR minima (such as, 20ms) time, upgrade OLD RRIAVG.The various criterions relevant to the behavior of current LV count value, nearest RRI difference and/or MEANVAR can be used alone or in combination to determine when to calculate the updated value of OLD RRI AVG.

When upgrading OLD RRI AVG, can be identical or different with the formula for the initial value calculating OLD RRI AVG after entering HV pattern for the formula calculating OLD RRI AVG.Such as, in both cases, can be averaging to calculate OLD RRIAVG to nearest 16 (or another quantity) RRI.In other embodiments, with after entering HV pattern for calculate OLD RRI AVG multiple RRI compared with, in response to resetting the nearest RRI that can use varying number when LV enumerator upgrades OLD RRIAVG during HV operator scheme.

If be greater than low variability threshold value in frame 310 current RRI difference but be not more than maximum variability threshold value (be negative decision at frame 312), then reduce in frame 316LV count value.Generally speaking, in response to the heart beating variability one by one of medium (moderate), LV count value reduces (frame 316).In response to high variability, it is zero (frame 314) that LV enumerator is reset.In response to low variability, LV count value increases (frame 320), as long as current MEANVAR is greater than predetermined minimum value.When RRI difference is less all the time, LV count value will reach sudden change threshold value.The evidence that this low variability of RRI that HR increases is the relevant rhythm of the heart causing the state that is transformed into 2 is attended by after the high variability (MEANVAR is larger than minimum threshold) of a time period.

Generally speaking, in response at least four kinds of situations, can there is never correlation behavior 1 and arrive the conversion of correlation behavior 2.Cause a kind of situation of the state that is transformed into 2 be as described in connection with figure 5 during LV operator scheme, detect that HR increases suddenly.Cause another situation being transformed into state 2 from state 1 be as described in connection of figure 8 during HV operator scheme, detect that being attended by the RRI variability increasing HR reduces suddenly.Cause the another situation of the state that is transformed into 2 to be medium reductions of RRI variability being attended by the evidence increasing HR as described in connection with figure 7, it causes the switching from the HV pattern state 1 to LV pattern.These situations relate to separately and are less than mutation rate limit value but the detection suddenlyd change to the rhythm of the heart of prescribing a time limit under being greater than verification and measurement ratio at the HR through estimating.When the HR through estimating is greater than mutation rate limit value, independent of HR mutating detector and RRI variability mutating detector, there is the conversion (as what see in the step 157 of Fig. 4) from state 1 to state 2.This higher HR is exactly separately the reason of the relevant rhythm of the heart.

To the sudden change of HR, the sudden change of RRI variability or being enough to of detecting all the time of heart beating causes individual count device described herein to reach predefine abrupt climatic change threshold level one by one time the RRI variability of the reduction arbitrary situation relevant with the combination of the HR of increase be referred to as " sudden change " of the rhythm of the heart in this article.Sudden change ensures to monitor to detect and distinguish tachycardia to the additivity 2 of EGM signal.Thus, the various criterions that the combination of the RRI variability of definable and the unexpected increase of HR, the unexpected reduction of RRI variability or the HR increased all the time and reduction is relevant, to use from state 1 to during the conversion of state 2 in realization.

Fig. 9 is for comparing NF HR and FF HR to verify the flow chart 350 of a kind of method of NF HR estimated value when meeting other abrupt climatic change criterions for being transformed into state 2 from state 1.Process shown in flow chart 350 may correspond to the frame 288 in the frame 243 or Fig. 7 of the frame 168 of Fig. 4, Fig. 5.

At frame 351, the enumerator for detecting sudden change during state 1 increases.This enumerator can be used for counting the outer enumerator of scope of (LV pattern) to the RRI dropped on beyond expection RRI scope or counting the LV enumerator of (HV pattern) for contrasting the little RRI difference of LV threshold value.If discontinuous counter value increases, then cause the current NF RRI increased at frame 352 buffer memory.The NF RRI that buffer memory is made contributions to abrupt climatic change is to use when determining the tolerance of NF HR.

Method for calculating the tolerance of NF HR according to the NF RRI of buffer memory depends on the abrupt climatic change threshold value of the outer enumerator of the scope of being applied to or LV enumerator.If abrupt climatic change threshold value is X, then at frame 356, the RRI of X-1 buffer memory is used to calculate NF HR estimated value.Value X can be nominally 10, but can such as change between 6 and 16.

In flow chart 350, NF HR estimated value is called as NF RR and changes interval and be measured as the average of two in the RRI of X-1 buffer memory.Such as, if change-detection threshold value is set as 10, then can be averaging to calculate NF RR to the minimum RRI of the 6th and the 7th in the RRI of nearest 9 buffer memorys and suddenlys change interval.

If meet sudden change threshold value (the outer count value of scope or LV count value meet or exceed sudden change threshold value) at frame 361, then FF EGM signal is used for verification computation is the NF HR estimated value that NF RR changes interval.As described earlier, if as frame 362 definitely FF EGM signal be insecure, then FF signal is not used in checking NF HR estimated value.In response to NF HR estimated value and abrupt climatic change, the conversion of state 2 can be occurred to.

If FF EGM signal is reliably (frame 362), then calculates FF RR at frame 363 and change the estimated value of interval as HR.Interval similar mode (that is, being averaging RRI selected by X-1 nearest FF RRI) can be changed with NF RR to calculate FF RR and change interval.These intervals, can be X-1 nearest continuous FF RRI and need not to be the FF RRI of the NF RRI corresponding to buffer memory in time.When the NF RRI that buffer memory causes change count value to increase, (intervening) the between two parties RRI not causing change count value to increase not buffer memory and be not used in and calculate NF RR and change interval.Thus, the NF RRI changing interval for calculating NF may not be continuous X-1 nearest interval.In an alternate embodiment, whenever buffer memory NF RRI, can the corresponding FF RRI of buffer memory to use when calculating FF RR and changing interval.

At frame 364, compare NF and FF and change interval to determine whether NF and FF heart rate estimated value roughly mates.If NF and FF changes interval and (such as, be about 20ms among each other) in predetermined threshold, then FF and NF heart rate estimated value is confirmed as roughly mating.Meet abrupt climatic change criterion, thus cause the state that is transformed into 2 at frame 370.

If roughly do not mated in frame 364NF and FF change interval, then detection algorithm remains in state 1.Process return of value frame 354 shown in Fig. 9 is to wait for the increase next time of discontinuous counter device.

Figure 10 is to provide the flow chart 400 of the general view of the operation performed during correlation behavior 2.At frame 402, tachycardia detection algorithm continues use and senses the EGM signal gathered while vector senses from two differences.Two sensing vectors detect operation for performing tachycardia, and these tachycardia detections operation is included in frame 403 and calculates tachycardia expection RRI scope, abandons noise/pseudomorphism at frame 404, perform overall signal morphological analysis at frame 405 and extract specific heart beating feature for carrying out analyzing adjuncts when needed at frame 407.

At frame 403, calculate tachycardia desired extent according to one of EGM signal (such as, NF EGM signal).Tachycardia desired extent is similar to the expection RRI scope used in state 1.But tachycardia desired extent represents according to the current relevant rhythm of the heart but not the RRI scope of previous normal uncorrelated rhythm of the heart expection.Similar with the expection RRI scope calculated during state 1, the tachycardia desired extent of frame 403 calculates to comprise and uses the weighted sum of current RRI and last RRMEAN and RRMAD value to calculate RRMEAN and RRMAD value.Initial value for calculating RRMEAN and RRMAD of tachycardia desired extent can be set to nominal value (such as, 500ms), and uses nearest five RRI (or RRI of another quantity) to be adjusted to actual value rapidly.

The last expection RRI scope calculated in state 1 before being transformed into state 2 is stored, and will not upgrade during state 2.Thus, the expection RRI scope of state 1 is frozen during state 2 operates.After being transformed into state 2 from state 1 exist expection RRI Range Representation return normal cardiac rhythm after expection HR.Thus, the expection RRI value range be transformed into after state 2 is stored to use when controlling to convert back state 1, as will be described in detail below.

At frame 404, perform noise/pseudomorphism and abandon process, each FF and NF EGM signal of this process analysis is to determine that can destroy tachycardia distinguishes the noise of algorithm or the existence of pseudomorphism.Imparting noise/pseudomorphism is classified to get rid of the impact of destroyed heart beating on tachycardia differentiating method by each heart beating.The various methods for detecting noise in EGM signal or pseudomorphism can be used.A kind of method current heartbeat being classified as go to pot heart beating or the heart beating that do not go to pot is described in conjunction with Figure 18 hereinafter.

At frame 405 and 407, each heart beating classifying as the heart beating that do not go to pot is analyzed on form ground as required.At frame 406, the result of configuration analysis and specific heart beating feature analysis is made contributions to VT evidence tolerance in the mode of accumulation on the basis of heart beating one by one.As described in detail in this article, according to the ad hoc rule relevant to the configuration analysis of FF EGM signal and/or NF EGM signal and/or the specific heart beating feature of FF and/or NF EGM signal, heart beating ground adjustment VT evidence enumerator one by one.

In some rhythms of the heart, compared with normal sinus rhythm heart beating, the change about the specific heart beating feature of FF or NF EGM signal can have assesses high tachycardia separating capacity than independent to the configuration of identical signal.Thus, specific heart beating feature is for strengthening sensitivity and the specificity of tachycardia differentiating method.

At frame 408, tachycardia desired extent and VT evidence count value are used to be sure of state 3 to detect VT and to proceed in VT testing process, or make the determination returning uncorrelated state 1.When making the judgement of return state 1, can use and operate from state 1 the expection RRI scope stored.If VT evidence count value reaches detection threshold, then detection VT is also occurred from correlation behavior 2 to the conversion be sure oing state 3.As being described in more detail below, criterion is defined as the conversion getting back to 1 between domination state 2 and state 3 and from state 2.This criterion can comprise and puts on noise/pseudomorphism and abandon the result of analysis 404, VT evidence accumulation (frame 406), tachycardia desired extent (frame 403) and the institute from state 1 and store and expect the requirement of RRI scope.

Figure 11 is the flow chart 410 for carrying out the method distinguished during state 2 operation between VT (can the treat) rhythm of the heart and SVT (untreatable) rhythm of the heart.Generally speaking, prescribe a time limit when the HR through estimating is greater than tachycardia verification and measurement ratio lower limit but is less than on SVT verification and measurement ratio, current cardiac frequency is classified as " the untreatable rhythm of the heart ", unless proved by the accumulation of VT evidence on the basis of heart beating one by one.At frame 412 and 414, two different EGM sensing vectors are used to record two different EGM signals (such as, FF and NF EGM signal as above) simultaneously.

In one embodiment, at frame 416, the NF signal sensed at frame 412 is used to sense ventricular cardiac event, i.e. R ripple.After frame 416 senses ventricular cardiac event, analysis window is set at frame 418, thus be defined for analyze sense the interval of FF and the NF EGM signal of heart beating.Although can be used for arranging analysis window based on NF sensed event NF EGM signal, for waveform morphology analysis and the object extracting specific heart beating feature, this analysis window can be applicable to NF signal and FF signal.In other embodiments, based on from corresponding EGM signal sensing to event, multiple independent analysis window can be set and be applied to EGM signal.

At frame 420, perform the morphological analysis to the FF EGM signal in analysis window.Owing to the form of the form of the whole EGM signal in analysis window with the EGM template obtained on similar time window being compared, therefore this analysis is called as " entirety " morphological analysis.In other words, when performing configuration and analyzing, do not isolate the particular amplitude in analysis window, slope or other times point and analyze.Waveform morphology as a whole during analysis window and known template form are as a whole compared to determine the matching degree between unknown heart beating and known template.Analysis window generally at least contains QRS wave group, but depends on the precision that analysis window persistent period and R ripple detect, and can comprise EGM signal more or less.

Variform parser is available, can apply these morphological analysis algorithms at frame 420.Generally speaking, perform morphological analysis, thus the configuration of the EGM signal sensed during the analysis window of unknown heart beating and the form of known heart beating are compared to use when sorting out this unknown heart beating.Such as, when wavelet analysis, the form of FF EGM signal and the known template of normal sinus rhythm heart beating are compared determine whether R ripple signal matches with normal sinus rhythm template.Can with reference to U.S. Patent No. 6,393,316 (people such as Gillberg), this patent by reference entirety is incorporated into this.

Form coupling mark is calculated as the configuration sensing EGM signal to the institute of unknown heart beating and the measurement of closely mating more than the configuration of the known template of normal conduction heart beating has at frame 422.High coupling mark generally indicates unknown sensing ventricular heartbeat to be the conduction heart beating being derived from atrium chamber.Low coupling mark generally indicates sensed ventricular heartbeat be not the heart beating of normal conduction and come from ventricle chamber.

In practice in the past, form mark and threshold value compare by VT detection algorithm, and correspondingly sort out sensed heart beating.Such as, in small echo morphological analysis, if VT heart beating threshold value can be configured to small echo coupling, mark drops under this threshold value, and unknown heart beating is classified as VT heart beating.If coupling mark exceedes threshold value, then SVT heart beating is sorted out in unknown heart beating.The classification based on threshold value based on this form coupling mark detects VT or SVT.

But in some cases, whole matching mark can drop to and is in close proximity to selected threshold value.Configuration coupling mark may be responsive not to the slight change of the EGM signal occurred between the VT heart beat period of some types.Thus, when being used to be separated VT and SVT heart beating based on configuration coupling mark, fixed threshold border, when heart beating is actually VT heart beating, SVT heart beating can be classified as, otherwise or.In clinical practice, there is the remarkable overlap of the form mark of VT and SVT heart beating, this can cause the undetected of VT or the flase drop to VT.

Thus, during state 2 operates, except the configuration coupling mark for accumulating VT evidence, use the analyzing adjuncts of specific heart beating feature.At frame 424, add heart beating feature from the EGM signal extraction of given heart beating.Analyze selected heart beating feature to improve sensitivity and the precision of SVT/VT differentiation.These heart beating features are referred to herein as " specific " heart beating feature, because compared with configuration mark, these features " more closely " observe EGM signal.Specific heart beating feature can be the isolation characteristic (such as, relevant to amplitude, slope or other waveform characteristics) of EGM waveform, and the particular point in time of these features in frame 418 is arranged analysis time window or sub-interval occur.In certain embodiments, if FF EGM signal is used to obtain configuration mark, then specific heart beating feature can comprise the feature than FF EGM signal space localization more.Such as, the configuration mark of the NF EGM signal obtained across the analysis window identical with FF configuration mark can be considered to heart beating feature more more specific than FF configuration mark, because compared with FFEGM signal, NF EGM signal is the signal of space localization more.Thus, specific heart beating feature be with from across selected analysis time window the whole persistent period EGM signal acquisition configuration mark compared with in time or the feature of EGM signal of spatially more isolating or localizing.

Based on the configuration coupling mark result calculated at frame 422, select the specific heart beating feature extracted at frame 424.The specific heart beating feature extracted at frame 424 may correspond in the U.S. Patent application S/N.12/415 at common transfer, and the beat morphology parameter of general description in 445, this patent application by reference entirety is incorporated into this.Heart beating feature can be extracted from NF EGM signal, FF EGM signal or both combinations.Rule is applied to configuration coupling mark and specific heart beating feature, thus frame 426 on the basis of heart beating one by one by increasing or reducing the evidence that VT evidence count value accumulates VT, as will be described in detail below.

If exceed in frame 428, VT evidence count value and can treat rhythm of the heart detection threshold, then occur to the conversion be sure oing state 3 at frame 430.In various embodiments, changing to carry out state 3 at frame 430, the additional criteria being applied to RRI data must be met.If VT proves that counting does not exceed detection threshold, then at frame 432, this process proceeds to next heart beating to continue sensing NF and FF EGM signal by being back to frame 412 and 414.

Figure 12 is the flow chart 450 of the illustrative methods extracting specific heart beating characteristic sum accumulation VT evidence on the basis in heart beating one by one.At frame 452, based on NF sensed event, the analysis window of current heartbeat is set.At frame 454, operating period inspection of the body statue and movements analytical method (such as wavelet analysis) analyzes FF EGM signal.The measurement of the FF FGM signal during form coupling mark can be calculated as analysis window and the whole matching between the morphological template of normal conduction heart beating (that is, sinus heart beating).At frame 454, calculate FF configuration coupling mark (FFMS).

At frame 460 to 466, FFMS is compared from the different confidence districts that tachycardia is distinguished.As shown in flow chart 450, multiple district can be defined as comprising SVT confidence district (frame 460), SVT grey area (frame 462), VT grey area (frame 464) and VT confidence district (frame 466).Suppose that the possible form coupling mark indicating the accurate match in the analytical method resolution between unknown heart beating and the storing template of normal conduction heart beating is that 100, SVT confidence district can be defined as any mark higher than 85.SVT grey area can be defined as being less than or equal to 85 but be more than or equal to the mark of 70.VT grey area can be defined as being less than 70 but be more than or equal to the mark of 40, and VT confidence district comprises any mark less than 40.Depend on used form matching algorithm, the clinical data relevant to the confidence level of form mark, clinician's preference or other factors, definable is for separating of other threshold values in FFMS district.

Depend on as decision box 460 to 466 definitely FFMS mark the district that falls into, at each frame 470,474,478 and 482, according to EGM signal measurement or specific heart beating feature can be calculated.The additional specific heart beating feature calculated is the specific heart beating feature needed for the various heart beating characterization rules of subsequent blocks 472,476,480 or 484 application.The additional heart beating feature that selection is extracted and the heart beating characterization rules mating mark district according to the form of current FFMS and apply strengthen sensitivity and/or the specificity of tachycardia detection algorithm.At frame 486, according to applying at frame 472 to 484 and being found to be genuine rule, FFMS and/or specific heart beating feature will be used to increase or reduce VT evidence count value.

At frame 486, in response to the result of one or more applied heart beating characterization rules, increase or reduce VT evidence count value.Based on one or more heart beating characterization rules of " exciting (fire) " (that is, being found to be very), on the basis of heart beating one by one, can effectively increase or reduce the impact that FFMS accumulates VT evidence.

If because the matched FFMS between unknown heart beating and normal heartbeat template falls into SVT confidence district, then extract additional SVT confidence district's heart beating feature at frame 470.The specific heart beating feature extracted applies those specific heart beating features needed for one or more SVT confidence district's heart beating characterization rules at frame 472.

One or more SVT confidence district's heart beating feature can be any feature gone out from FF and/or the NFEGM signal extraction during analysis window, and can comprise the form determined with known comparing of NF EGM template according to the NF EGM signal of current unknown heart beating and mate mark.

The SVT confidence district's heart beating characterization rules applied at frame 472 can comprise threshold value relatively or other criterions one or more extracted heart beating feature be associated with SVT heart beating or VT heart beating.In one embodiment, one or more SVT confidence district's heart beating characterization rules be defined as mark may with the evidence of the discovery conflicting VT of FFMS in SVT confidence district.If meet this rule, then correspondingly adjust VT evidence enumerator at frame 486.

Figure 13 A is the flow chart 500 of application SVT confidence district's heart beating characterization rules.In one embodiment, no matter when FFMS falls into SVT confidence district is all applied for using NF EGM signal to detect the rule of the evidence of exception (VT) heart beating.Thus, when FFMS falls into SVT confidence district (frame 502), NF form coupling mark (NFMS) of the NF EGM signal in frame 504 computational analysis window.Additionally or alternatively, at frame 504, can calculate or measure the specific heart beating feature of EGM signal.Any example of specific heart beating feature described herein uses when all can apply NF abnormal heartbeats rule in SVT confidence district.

At frame 506, NFMS and VT and SVT detection zone are compared, this VT with SVT detection zone can define identical with FFMS detection zone or similarly.If NFMS falls into the district corresponding with VT, then this discovery is the evidence of abnormal heartbeats.The result that this discovery and overall FFMS based on " more closely observing " heart signal fall into SVT confidence district contradicts.In other embodiments, the specific heart beating feature of FF and/or NF signal can be extracted and compare the normal stencil value of itself and each feature to detect the evidence of VT heart beating.

Frame 518 state NF abnormal heartbeats rule be true before, for noise/pseudomorphism destructive inspection FF and/or NF EGM signal.The result that noise/pseudomorphism abandons process (frame 512) is used to determine whether FF and NF EGM signal is not destroyed signal at decision box 510 and 514 respectively.Multiple noise/pseudomorphism detection algorithm can be used for the noise or the pseudomorphism that heart beating or EGM band (strip) are isolated into pollution.In one embodiment, if be found to be at frame 510FF signal the signal that goes to pot, then current heartbeat is skipped at frame 516.Do not adjust the VT evidence count value of current heartbeat.

But if FF signal is found to be the signal that goes to pot (frame 514) for the signal NF signal that do not go to pot, then state that NF abnormal heartbeats characterization rules is false at frame 508.This NF signal is not used in confirmation or overthrows FFMS result.If FF and NF EGM signal is not destroyed by noise or pseudomorphism, be then found to be very in this rule of frame 518.This result is used to adjustment VT evidence count value in conjunction with FFMS result.

Refer again to Figure 12, at frame 486, based on NF abnormal heartbeats rule and the result of FFMS, adjust VT evidence count value.Generally speaking, in response to only based on the strong evidence of the SVT heart beating of FFMS, VT evidence count value will reduce.But, the VT evidence count value size of successively decreasing may than in NF abnormal heartbeats rule for true time is less.It is only based on the confidence level of the SVT heart beating of FFMS that the evidence of the abnormal heartbeats in NF EGM signal reduces unknown heart beating.

In an illustrative embodiment, when FF EGM form mark falls into SVT confidence district, at frame 486, can reduce according to following process VT evidence count value:

If NF abnormal heartbeats rule is true, then VT evidence reduces 0.5

Otherwise VT evidence reduces 2.0

Terminate

Because FF EGM mark is high enough to fall into SVT confidence district, therefore VT evidence count value reduces.But when NF abnormal heartbeats rule is found to be true and FF and NF signal all not by sound pollution, VT evidence count value reduces less amount.

If NFMS or other specific heart beating features one or more are found to be " normally " (that is, corresponding with coming from supraventricular heart beating), then NF EGM furanone FFMS result.NF abnormal heartbeats rule may be found to be vacation, and based on this evidence, VT evidence count value can reduce (or remaining on null value) from currency.But if specific heart beating is characterized as "abnormal" (that is, may indicate the heart beating coming from ventricle), then this evidence contradicts with FFMS and significantly reduces the confidence level that this heart beating is SVT heart beating.In the case, VT evidence count value still can reduce, but with to compare decrement when with configuration mark, specific heart beating characteristic results both supports that current heartbeat is the common discovery of SVT heart beating less.

If FFMS falls into SVT grey area (as determined at frame 462), then calculate SVT grey area heart beating feature at frame 474.In the case, extract and can provide the specific heart beating feature of the evidence of VT heart beating, evidence and this heart beating of this VT heart beating are that the FFMS result of SVT heart beating contradicts.The evidence of this contradiction can comprise following in one or combination: the larger difference of very low NFMS, specific NF heart beating feature compared with corresponding normal mode plate features and/or specific FF heart beating are relative to the larger difference of normal mode plate features.

Heart beating feature in SVT grey area can comprise the special characteristic of NF and/or FF signal.In one embodiment, the SVT grey area heart beating feature extracted at frame 474 comprises greatest gradient, R wave width, R ripple symmetry index (variability of the uplink and downlink of R ripple) and QR index (ratio of Q wave-amplitude and R wave-amplitude).Other features that can comprise in various embodiments comprise the time difference between R ripple polarity concordance between FF and NF signal and/or template R ripple polarity, the time difference of NF and FF peak amplitude and NF and FF greatest gradient.Will be appreciated that, multiple specific heart beating feature can be selected.Selection is found in the heart beating feature between SVT and VT heart beating with highest region point ability, for when application rule in order to increase on the basis of heart beating one by one or to reduce VT evidence count value.Can ignore and not find that improving tachycardia distinguishes the sensitivity of algorithm and specific feature.

At frame 476, application rule is improved and is distinguished algorithm to the sensitivity of VT.These rules, for the specific heart beating feature of examination of evidence of VT heart beating, at frame 486, fall into these specific heart beating features of SVT grey area based on FFMS result and VT evidence count value will be made to increase and non-reduced.

Figure 13 B applies the flow chart 520 of a kind of method of VT heartbeat when being for falling into SVT grey area at FFMS.If FFMS falls into SVT grey area (frame 522), then check that the noise/pseudomorphism of FF signal abandons result (frame 524) at frame 526.If FF signal goes to pot, then do not perform the analyzing adjuncts to specific heart beating feature.Skip whole heart beating at frame 527, and any adjustment is not made to VT evidence enumerator.

If FF signal does not go to pot, then calculate FF greatest gradient at frame 528.If the absolute value of the greatest gradient of analysis window period FF EGM signal is less than low slope threshold value (decision box 538), then do not perform analyzing adjuncts.When the greatest gradient of FF EGM signal is lower than threshold value, FF EGM signal can not have enough signal intensitys to assess specific heart beating feature.This VT heartbeat is false (frame 534).

If FF signal does not go to pot and meets minimum slope requirement, then analyze additional specific heart beating feature to detect the evidence of VT in current heartbeat.One of NF and FF signal or both can be used for perform analyzing adjuncts to detect the evidence of VT.

Such as, if be confirmed as being not destroyed (frame 532) based on input (frame 524) the NF signal carrying out self noise/pseudomorphism discard algorithm, then NFMS can be calculated at frame 536.If NFMS falls into VT confidence district (that is, the extremely low coupling between NF EGM signal and known normal heartbeat template), then it is true for exciting at frame 544, VT heartbeat.In current heartbeat, this evidence of VT will use when adjusting VT evidence count value.

If NF EGM signal goes to pot (frame 532) if or NF configuration mark be greater than VT confidence district (frame 538), then perform the analyzing adjuncts of the specific heart beating feature from FF signal extraction to detect the possible evidence of VT in current heartbeat.Such as, can measure or calculate the specific heart beating feature of FF at frame 540, and at frame 542, itself and each the specific heart beating feature measured according to FF normal heartbeat template or calculate be compared.If any FF specific heart beating feature is different from template heart beating characteristic remarkable ground, then in current heartbeat, detect the evidence of VT.

In one embodiment, the FF ratio (in this article also referred to as QR index) of FF R wave width, FF R ripple symmetry and Q wave-amplitude and R wave-amplitude is compared with the character pair of FF normal heartbeat template respectively.Threshold difference between any one FF specific heart beating characteristic sum FF normal heartbeat template will cause VT heartbeat to excite at frame 544.

Generally speaking, if FF signal goes to pot, then skip whole heart beating and adjust for VT evidence.But if FF signal does not go to pot do not meet minimum slope requirement, then do not perform the analyzing adjuncts of the evidence for detecting VT.At the frame 486 of Figure 12, adjust VT evidence tolerance based on FFMS.But, if FF signal does not go to pot and meets minimum slope requirement, then check specific heart beating feature.Selected heart beating feature can be knownly to change between VT heart beat period, but may not yet be enough to make whole FFMS fall into any feature in VT district.If FF heart beating feature is undiscovered meet VT evidence criterion, then NF EGM signal still can be used to the evidence detecting VT in current heartbeat.If NF signal does not go to pot and causes the NFMS in VT confidence district, then, when adjusting VT evidence count value, this VT evidence will use for FFMS result.

In an illustrative embodiment, when overall NFMS is very low, not from the specific heart beating feature of NF EGM signal extraction for checking the evidence of VT.But, should be appreciated that in an alternate embodiment, can for the specific heart beating feature of the examination of evidence of VT from NF EGM signal.Such as, if NFMS is higher than VT confidence district but still in grey area, then check that specific NF heart beating feature is to identify the feature of instruction VT.In addition, if NFMS falls into SVT district (Lycoperdon polymorphum Vitt or confidence) thus confirmation FFMS result, then the process shown in Figure 13 B can directly proceed to frame 534, thus adjusts VT evidence count value based on FFMS, and without the need to the further analysis to specific FF heart beating feature.

Referring again to the frame 486 of Figure 12, adjusting VT evidence count value according to whether meeting VT heartbeat at frame 476.If VT heartbeat is true, then increase a pre-determined incremental amount in frame 486, VT evidence count value.If VT Evidence Rules is not true (that is, based on the evidence to the no abnormal heart beating of the additional examination of FF and/or NF signal), then because FFMS falls into SVT grey area, VT evidence counting reduces.In the case, VT evidence counting reduction one decrement, this decrement is less than when falling into SVT confidence district at FFMS and do not have the evidence of abnormal heartbeats at NF signal for reducing the decrement of VT evidence tolerance.

When FF configuration mark falls into SVT grey area, following process can be used to adjustment VT evidence counting:

If VT heartbeat is true, then VT evidence increases 0.625

Otherwise VT evidence reduces 0.5

Terminate

Mentioned herein be applied to the incremental change of VT evidence counting and the particular value of decrement is illustrative, and be adjustable to provide the optimum sensitivity and specificity of distinguishing algorithm.In addition, will be appreciated that, the various deformation of specific heart beating feature and to substitute and combination can when FFMS falls into SVT grey area for detecting the evidence of VT.

If FFMS falls into VT grey area (frame 464), then calculate at frame 478 and apply specific heart beating feature needed for VT grey area heart beating characterization rules at frame 480.Extract additional specific heart beating feature, it may be more the evidence of normal conduction (SVT) heart beating but not VT heart beating that these additional specific heart beating features provide with the conflicting current heartbeat of FFMS result.Additionally or alternatively, can execution analysis to detect the supporting evidence that current heartbeat is very likely the VT heart beating supporting FFMS result.

Specific heart beating feature can be calculated according to FF and/or NF EGM signal and the corresponding template feature of itself and normal heartbeat is compared.In one embodiment, at frame 480, application FF normal heartbeat rule detects the evidence of the normal heartbeat feature in FF signal, and this evidence can contradict with the FF configuration mark in VT grey area.At frame 480, also Second Rule can be applied (namely, NF normal heartbeat rule) detect evidence or the configuration of the normal heartbeat feature in NF signal, it can mate mark with FF configuration and contradict and the follow-up adjustment affecting VT evidence count value.

What calculate at frame 478 can comprise in the frame 480 heart beating feature applied needed for normal heartbeat rule any feature listed herein, includes but not limited to: the ratio of greatest gradient, R ripple polarity, R wave width, R ripple symmetry index, Q ripple and R wave-amplitude, be measured as the amplitude peak skew of the time difference between FF signal and the absolute peak swing of NF signal and be measured as the slope peakdeviation of the time difference between the absolute greatest gradient of FF signal and the absolute greatest gradient of NF signal.

Figure 13 C is the flow chart of a kind of method for applying VT grey area rule.When FFMS falls into VT grey area (frame 522), check that the result of noise/pseudomorphism discard algorithm (frame 554) is to determine whether FF signal goes to pot (at frame 556).If FF signal goes to pot, then as indicated at frame 558, skip whole heart beating when not adjusting VT evidence count value.

If FF signal does not go to pot, then the FF greatest gradient between frame 560 computational analysis window phase, and at frame 562, itself and low slope threshold value are compared.If FF greatest gradient is less than low slope threshold value, then can not perform the further analysis to FF specific heart beating feature.FF normal heartbeat rule is false (frame 566).The process of application VT grey area rule advances to frame 572 to apply NF normal heartbeat rule, as described further below.

If FF greatest gradient exceedes low slope threshold value (frame 562), then calculate the specific heart beating feature of FF at frame 564.FF specific heart beating feature can comprise any heart beating feature previously listed in this article.The specific heart beating feature that uses of FF normal heartbeat rule is selected as providing current heartbeat may be SVT heart beating but not those specific heart beating features of the evidence of VT heart beating of implying of FFMS.At frame 568, specific heart beating feature is compared to the corresponding heart beating feature of FF normal heartbeat template.

FF normal heartbeat feature can require that one of two or more characteristic heart beating features or combination fall into the scope of predefine threshold value or corresponding FF normal mode plate features, thus the evidence of statement SVT heart beating.In one embodiment, if FF QR index, FF R wave width and FF R ripple symmetry index are all in the corresponding scope of corresponding FF normal mode plate features, then at frame 570, statement FF normal heartbeat rule is true.If the specific heart beating feature of any one FF does not meet normal heartbeat requirement, then FF normal heartbeat rule is false (frame 566).In other embodiments, FF normal heartbeat rule can comprise " OR (or) " operator, if at least one or some specific heart beating features are found to be roughly mate with normal heartbeat template, then should this rule be met " OR " operator.

When meeting FF normal heartbeat rule, exist with from falling into the FF configuration mark of VT grey area and providing the conflicting evidence of specific heart beating feature of the evidence of SVT.The adjustment to VT evidence count value that this contradiction evidence will affect at the frame 486 of Figure 12, as will be described hereinafter.

Except FF normal heartbeat rule, NF normal heartbeat rule can be applied to check the evidence of SVT further in NF EGM signal.Thus, after determining that FF normal heartbeat rule is true or false (frame 566 or 570), the result (frame 554) of noise/pseudomorphism discard algorithm is used to check at frame 572, NF signal is to determine whether this signal goes to pot.If gone to pot, then do not perform the further analysis to NF signal.NF normal heartbeat rule is false (frame 574).It is reliable that NF signal is not considered to detecting the evidence of the SVT heart beating with FFMS contradiction.

If NF signal does not go to pot, then calculate NFMS at frame 576.If NFMS very high (such as, as determine at frame 578 in SVT confidence district), then NF normal heartbeat rule excites is true (frame 580).High NFMS is detected as the evidence of the normal conduction heart beating in the Shang Shi district coming from heart.In other embodiments, NF normal heartbeat rule can comprise comparing of the individual features of NF specific heart beating feature and the normal template of NF.

Refer again to Figure 12, at frame 486, according to the result of application VT grey area rule, adjustment VT evidence enumerator.If FF normal heartbeat rule and NF normal heartbeat rule are not true, then fall into VT grey area in response to FFMS, VT evidence count value increases a pre-determined incremental amount.If FF signal goes to pot, then skip whole heart beating and adjust for VT evidence.

If FF normal heartbeat rule or NF normal heartbeat rule in one or two be found to be very, then depend on the strength of confirmation of SVT heart beating, VT evidence count value can increase less incremental change or a reduction.In an example of the process of VT evidence count value that adjusts after the rule of application VT grey area of frame 486 use be:

If FF normal heartbeat rule is true, then VT evidence reduces 0.375

NF normal heartbeat rule is true else if, then VT evidence reduces 0.5

Otherwise VT evidence increases 0.75.

In this example, only have FF and NF normal heartbeat rule to be false, VT evidence just increases.If FF normal heartbeat rule is true, then the evidence of SVT heart beating owing to finding in the specific heart beating feature of FF, VT evidence count value reduces a relatively little decrement.Compared with falling into the FFMS of VT grey area when the VT evidence of accumulation current heartbeat, prove that the FF of SVT heart beating specific heart beating feature is endowed larger weight.

If NF normal heartbeat rule is true, then VT evidence count value reduces slightly to be greater than the decrement that FF normal heartbeat rule is true time.The NF signal evidence of SVT heart beating is considered to the SVT heart beating evidence stronger than the FF of SVT heart beating specific heart beating feature evidence, and the correct beat classification evidence stronger in VT grey area than FFMS.Will be appreciated that, in various embodiments, specific incremental change and decrement can give different values.

For current heartbeat, VT evidence tolerance is only adjusted once.In certain embodiments, multiple rule can be applied, and can based on think the single rule in the source correctly identifying current heartbeat with maximum confidence basis on make adjustment.In other embodiments, these rules can hierarchical mode be applied.The first rule excited is for determining the incremental change or the decrement that are applied to VT evidence enumerator.The highest rule excited is considered to have maximum confidence level in differentiation VT and SVT.In other embodiments, each incremental change of being associated with the multiple rules excited for given heart beating or decrement sum can be confirmed as to the clean adjustment of VT evidence count value.

If as frame 466 definitely FFMS fall into VT confidence district, then can on-demand computing VT confidence district's heart beating feature at frame 482, thus apply VT confidence district rule at frame 484.In the case, any rule of applying at frame 484 of definable with detect can with the possible evidence of the discovery conflicting SVT heart beating based on FFMS.Such as, the specific heart beating feature of FF and/or NF can be calculated, thus in VT confidence district, applying normal heartbeat rule, to detect this heart beating of instruction be the evidence of the SVT heart beating of normal conduction.Fall into SVT confidence district in response to FFMS, VT evidence count value can increase.But if normal heartbeat rule excites is true based on one or more characteristic heart beating feature (comprising NFMS or any other described in this article feature), then the incremental change being applied to VT evidence enumerator can reduce.

In other embodiments, when FFMS fall into confidence district, SVT confidence district or VT confidence district time, at frame 486, directly can adjust VT evidence count value.Such as, if FFMS very low (that is, in VT confidence district), then one (or another incremental change) can be added immediately in frame 486, VT evidence count value, and without the need to extracting specific heart beating characteristic sum application heart beating characterization rules.Due to the more high confidence level of VT heart beating, large when the incremental change being applied to VT evidence count value in the case falls into VT grey area than FFMS.Equally, when FFMS is very high and when falling into SVT confidence district, VT evidence count value can reduce immediately, and without the need to analyzing further heart beating feature.Relate to the rule improving sensitivity and specific specific heart beating feature and can be applicable to " Lycoperdon polymorphum Vitt " district, be somebody's turn to do the intermediate range that the possible range of configuration mark is only contained in " Lycoperdon polymorphum Vitt " district.

At frame 488, VT evidence counting is compared to detect VT with threshold value.If VT evidence counting reaches detection threshold, then occur to the conversion be sure oing state 3 at frame 490.If not yet reach detection threshold, then this algorithm proceeds to next heart beating at frame 492, and is back to frame 452 to continue accumulation VT evidence, as long as meet other criterions for maintaining in state 2.

Figure 14 is the flow chart 600 for the method across multiple FFMS grey area application rule.As the additional or replacement of district's dedicated rules that the frame 472,476,480 and 484 at Figure 12 is applied, other rules can be applied across multiple form mark district.Such as, the rule applied across whole grey area (that is, SVT grey area and VT grey area), whole VT district (that is, VT grey area and VT confidence district) or whole SVT district (that is, SVT grey area and SVT confidence district) of definable.These rules can check the specific heart beating feature of the strong evidence being considered to SVT or VT heart beating characteristic.If these rules are found to be very, then when adjusting VT evidence count value, these rules can surmount (override) any single district rule.

If as indicate at frame 602 FFMS fall into SVT or VT grey area any one (whole grey area), then perform the particular condition that additional EGM signal analysis is initiated by the ventricular extra-systolic caused tachycardia heart rate (PVC) to detect VT.The details relevant to the detection that heart rate starts, at U.S. Patent application S/N.12/430, is described in 301, and this patent application by reference entirety is incorporated into this.In brief, check that variability and this n the relative change recently between a RRI and front n RRI sum of n nearest RRI start to detect tachycardia heart rate.When variability be less than variability threshold value and relatively change be greater than relative change threshold time, current heartbeat is detected as tachycardia heart rate and starts heart beating.

Initiating heart beating can be the heart beating starting the Zao some heart beatings (such as, about 4 to 6 heart beatings) of heart beating than tachycardia heart rate.If starting the Zao heart beating of heart beating than tachycardia heart rate is PVC, then tachycardia is very likely the VT initiated by PVC.Thus, for any FFMS falling into whole grey area, the strict heart rate of application PVC starts rule to strengthen the sensitivity of the algorithm for detecting VT.

If at frame 608, such as, according to the method described in common ' 301 U.S. Patent applications transferred the possession of, detect tachycardia heart rate for current heartbeat and start, then for the flag check previous heartbeats that previous heartbeats can be indicated to come from ventricle.Single previous heartbeats can be checked, the 4th such as Zao than current RRI heart beating.In an alternate embodiment, for having there is the evidence of initiating PVC in the some heart beatings started at tachycardia heart rate before detection, one or more previous heartbeats can have been checked.

At frame 610 to 614, the multiple specific heart beating feature of previous heartbeats and each heart beating feature of normal template are compared.In an illustrative embodiment, need all these heart beating features significantly different from normal heartbeat template, thus determine that this heart beating has as PVC, the high likelihood of initiating tachycardia heart rate.

Frame 610 to 614 check specific heart beating feature in various embodiments between can be different.In flow chart 600, at frame 610,612 and 614, R ripple polarity, slope peakdeviation, R wave width and QR index and normal template are compared.If all these character representations depart from the change of normal template, then it is true for starting that rule excites at the strict heart rate of frame 620, PVC.If any one FF specific heart beating feature is significantly not different from normal template, then PVC initiate VT FF specific heart beating feature evidence not yet by force to other VT or the SVT evidences being enough to surmount overall FFMS result and district's dedicated rules and providing.

But, if do not provide the evidence of VT in the specific heart beating feature of frame 610 to 614, FF, then can additionally check that NFMS is to detect the evidence of initiating PVC at frame 616.When NFMS and FFMS of previous heartbeats both falls into VT confidence district, the strict heart rate of PVC will be met and start rule (frame 620).If FF specific heart beating feature (frame 610 to 614) and NFMS and FFMS (frame 616) do not meet regular criterion, then it is false for starting rule at the strict heart rate of frame 606, PVC.

If it is true for starting that rule excites at the strict heart rate of frame 620PVC, then, when adjusting VT evidence count value, this result can surmount other rules being applied to single FFMS district.In one embodiment, it is true for starting rule in response to the strict heart rate of PVC, increases to maximum (such as, 8) in frame 622, VT evidence count value.If it is false that the strict heart rate of PVC starts rule, then FFMS and one or more corresponding districts dedicated rules are for determining the adjustment to VT evidence count value for current heartbeat.

Figure 15 is the flow chart 625 of the process for applying the rule for detecting rhythm of the heart breakpoint, and this rule can across the application of multiple FFMS districts.Tachycardia detection method based on interval needs the RRI of the specified quantity in previous multiple nearest RRI (such as, 18 intervals interim between 24) shorter than tachycardia detection interval usually, so that detect tachycardia.This allows nearest RRI in a small amount longer and still suitably detect tachycardia when deficient sensing exists than tachycardia detection interval.But in some cases, long RRI can be the actual pause in the rhythm of the heart, instead of owe the result of sensing.When detecting interruption or the spontaneous termination of the tachycardia rhythm of the heart, detecting that actual (tube) length pauses is useful information.

Two the continuous normal heartbeat growing the interruption paused or in the labelling tachycardia rhythm of the heart can occur during atrial fibrillation, recurrent non-continuous VT or other untreatable rhythms of the heart.The relevant rhythm of the heart be T ripple cross the result of sensing time, also can detect two continuous normal heartbeat or long to pause.Thus, expect actual (tube) length RRI in detection two successive heartbeats and/or normal R waveform state, thus rhythm of the heart breakpoint can be identified and use it for avoid continuing to detect can treat tachycardia and the rhythm of the heart be non-continuous or the untreatable rhythm of the heart time treatment transmission.

Usually, at frame 626, any time during detection algorithm long RRI can be detected.Criterion for detecting long RRI can average based on the nearest RRI of a percentage ratio of previous RRI and/or specified quantity.Such as, if current RRI grow to few 25% than previous RRI and than average, the maximum of nearest 8 RRI (or nearest RRI of another quantity) or another tolerance long, then long RRI detected.Requirement that is additional or that replace can be that RRI is longer than minimum threshold interval.Various criterion can be used for detecting long RRI, but usually will comprise the comparing of tolerance with previous RRI and/or nearest RRI.

At frame 628 and 632, optionally perform and the analyzing adjuncts of one or more EGM signal are detected to the evidence of actual (tube) length RRI to detect support or long RRI can be indicated by the evidence owing to sense the deficient sensing caused.At frame 628, for the evidence of the deficient sensing R ripple occurred during long RRI, check one or more EGM signal.The evidence owing sensing R ripple can comprise the maximum conversion ratio exceeding threshold value, the peak swing exceeding threshold value and/or occur during long RRI can be corresponding with QRS wave group other signal characteristics.Such as, if signal amplitude exceedes the predetermined percentage of the R wave-amplitude sensed recently, then R ripple by deficient sensing, thus may cause the flase drop to long RRI between two parties.In particular example, if find the amplitude than current R wave-amplitude about 25% during the long RRI detected, then can exist and owe sensing R ripple.Can from the EGM signal detection identical or different with the EGM signal for detecting long RRI or other signal aspect evidences obtaining amplitude, conversion ratio or deficient sensed event.

Additionally or alternatively, can check that another EGM senses vector at frame 628, the sensed event of vector does not occur to sense about another to verify during the long RRI detected.Such as, if the long RRI about NF EGM signal detected, then check that FF EGM signal is to verify the sensed event do not occurred during the long RRI about NF signal about FF EGM signal.The FF sensed event occurred within interval short period relative to NF sensed event may correspond in identical heart beating.Thus, for interval that the interval about NF EGM signal check of the evidence detecting the deficient sensing R ripple about NF EGM signal can be defined as the shortening in the long RRI of NF or narrow.Such as, for FF sensed event, can check that after previous NF sensed event about 50 to 80ms started and in the interval terminated than Zao about 50 to the 80ms of current NF sensed event.If there is not FF sensed event during this interval, then long RRI may be that the actual (tube) length that can be detected as rhythm of the heart breakpoint pauses.

In a similar fashion, by checking about the long RRI whether sensed event occurring verify in the corresponding shortening interval of NF EGM signal about FF EGM signal measurement.The variant searching for the method for sensed event during shortening interval will determine whether sensed event occurs about two different EGM signals with the 1:1 ratio containing the interval of the one or more RRI about an EGM signal.Such as, if there are three events about FF signal during containing the interval about the single RRI of NF signal, then the deficient sensing about NF signal can be suspected.

If such as based on high-conversion rate, high amplitude, during long RRI, there is the evidence owing to sense R ripple about the sensed event of another sensing vector or other signal characteristics during long RRI, then rhythm of the heart breakpoint do not detected and tachycardia detection algorithm continues at frame 630.Any imminent treatment can be continued.If there is no owe the sensing R evidence of ripple, then can analyze sensed event form to verify that the R wave table terminating long RRI shows normal R waveform state at frame 632.Can calculate form coupling mark, and at frame 634, the high coupling mark terminating the R ripple of long RRI is detected as rhythm of the heart breakpoint.

The form coupling mark detected needed for rhythm of the heart breakpoint can be dependent on the length of long RRI.The coupling of the threshold value form needed for the rhythm of the heart breakpoint mark detecting different range RRI can be stored in tracing table, or can be defined as mating the exponential relationship between mark and RRI or other relations.In order to illustrate, if RRI is just than average slightly long (such as, the highest be about 25%) of previous RRI or RRI recently, then very high form coupling mark may be needed to detect rhythm of the heart breakpoint (such as, in SVT confidence district).If long RRI is than on average much longer (such as, being about more than 50%) of previous RRI or nearest RRI, then lower coupling mark (mark such as, in SVT grey area) can be accepted as the evidence of rhythm of the heart breakpoint.

In a particular embodiment, if current RRI grows to few 350ms than previous RRI or nearest 8 RRI on average large at least about 25%, then rhythm of the heart breakpoint detected when the form coupling mark of the QRS signal terminating long RRI is greater than 35/RRI, wherein RRI be current RRI in seconds and the scope of form coupling mark for from 0 to 100.

It is conceivable that, when detecting long RRI to verify rhythm of the heart breakpoint, can perform in frame 628 and 632 one or two.At frame 640, provide the response to rhythm of the heart breakpoint.To detect the response of rhythm of the heart breakpoint can comprise stop treatment, postpone treatment, the early treatment initiation menu for the treatment of reset in menu order, other evidences removing counting or VT, amplification SVT evidence, change current cardiac frequency classification or terminate the measurement of rhythm of the heart outbreak (episode) persistent period and reset outbreak timer.

Can occur during any state comprising the detection algorithm be sure oing state 3 the long detection paused, and can change between each state as required the long response paused detected as rhythm of the heart breakpoint.

Operate about tachycardia detection algorithm state 2, detect in response to pause, VT evidence count value can be set to zero.If overrun between stage of attack at current cardiac transmitted any treatment, then next treatment be ranked will be reset as first for the treatment of menu by programming is treated.Thus, next treatment transmitted after again arriving state 4 will be the first treatment of selected treatment menu, instead of proceed to more radical treatment.

The method of interrupting for detecting the rhythm of the heart can realize in any tachycardia detection algorithm, and is not limited to use in rule-based detection algorithm described in this article.The long RRI of monitoring can be performed and for deficient sensing evidence and/or the process of the long RRI of evidence analysis of normal heartbeat stopping long RRI, thus detect rhythm of the heart breakpoint in any tachycardia detection algorithm.

Figure 16 is the flow chart 650 of a kind of method for applying the rule for detecting rhythm of the heart breakpoint across whole SVT district.In this example, rhythm of the heart breakpoint rule is applied when FFMS falls into SVT confidence district or SVT grey area.Application rhythm of the heart breakpoint rule detects any heart beating normally with high confidence level.To pause with long or normal conduction heart beating that two continuous normal heartbeat are associated can be the designator of the pause in the tachycardia rhythm of the heart.

If the FFMS of current heartbeat falls into whole SVT district (frame 652), then check previous FFMS at frame 654.If the FFMS of current and previous continuous sensed event in SVT confidence district, then calculates NFMS at frame 655.As long as be greater than zero (or another predetermined threshold) at the NFMS of frame 656 current heartbeat, the evidence that two continuous FFMS in SVT confidence district just provide the rhythm of the heart to interrupt.At frame 670, it is true that rhythm of the heart breakpoint rule excites.

If current and previous FFMS does not fall into SVT confidence district (that is, one or two falls into SVT grey area or lower), then analyze current RRI to detect the long pause in the tachycardia rhythm of the heart.Generally speaking, if current RRI is confirmed as longer than previous RRI, then current RRI can represent long pause and rhythm of the heart breakpoint.In one embodiment, at frame 658, the moving average of current RRI and nearest RRI (such as, nearest eight RRI) is compared.Need current RRI longer than any factor of RRI moving average or RRI moving average.Such as, the weighter factor " M " in frame 658 can equal 1 or can be higher value.

If current RRI is longer or required percentage ratio moving average is long than moving average, then at frame 662, also current RRI and nearest previous RRI can be compared to verify long pause.Need current RRI than previous RRI length one predetermined factor or percentage ratio.Weighter factor " N " in frame 662 can equal such as about 1.3 or another selected value larger than 1.If the current RRI of negative decision based on the arbitrary comparison at frame 658 and 662 can not be long pause, then do not meet rhythm of the heart breakpoint rule and state that this rhythm of the heart breakpoint rule is false at frame 660.

If at frame 666, current RRI longer than previous RRI (such as, long than previous RRI at least about 30%), then check that whether FF specific heart beating feature is to determine to stop the heart beating of long RRI for normal heartbeat.At frame 666, can check that each FF specific heart beating feature is to determine whether specific heart beating feature closely mates with the respective specific heart beating feature of normal heartbeat template.In one embodiment, FF R ripple symmetry index, FF R wave width and FF QR index are compared with the character pair of FF normal heartbeat template respectively.If what each feature in these features all fell into normal heartbeat template can acceptance threshold scope, then meet rhythm of the heart breakpoint rule and state that this rhythm of the heart breakpoint rule is true at frame 670.The combined evidence hint current heartbeat of the specific FF heart beating feature evidence of the FFMS in long RRI, whole SVT district and the R ripple of normal conduction is rhythm of the heart breakpoint.Should be appreciated that in various embodiments, FFMS, FF specific heart beating feature, NFMS and/or NF specific heart beating feature can be used alone or in combination to determine that the end R ripple of long RRI is very likely normal heartbeat.

If rhythm of the heart breakpoint rule is true, then when adjusting VT evidence count value, this result can surmount other single district rule and FFMS.In one embodiment, at frame 672, be true in response to rhythm of the heart breakpoint rule, by the counter O reset of VT evidence.

If comprise more than one multi-region rule at tachycardia detection algorithm, then can perform these multi-region rules by hierarchical order, thus excite the rule for genuine highest level to surmount the result of more low-level rule to cause the adjustment to VT evidence count value.Multi-region rule can be write to make it mutually repel, namely there is not the probability simultaneously exciting more than one multi-region rule.Such as, whole grey area rule and whole SVT district rule can be write, thus two rules cannot be met for current RRI.In other embodiments, only exciting in single multi-region rule be true and every other multi-region rule is fictitious time, can respond and excite as really to the adjustment of VT evidence count value in multi-region rule.In the case, if it is true that more than one multi-region rule excites, then these multi-region rules can be considered to indecisive, and rely on the result of single district rule and FFMS to adjust VT evidence count value.

Alternatively, if multiple rules that can comprise single district and/or multi-region rule are applied to specific FFMS district, then these rules can predefine order be applied, thus only have the rule of all higher levels to be the more low-level rule of false just application.Once excite for given FFMS district rule, just adjustment VT evidence count value.

Figure 17 is for adjusting VT evidence count value in response to FFMS district rule on the basis of heart beating one by one thus cumulative evidence can treat the flow chart 800 of tachycardic process to detect.Process shown in flow chart 800 determines the suitable incremental change or the decrement that are applied to VT evidence enumerator for given heart beating in response to one or more applied heart beating characterization rules.Not shown can be applicable to determines whether should adjust VT evidence count value and other criterions of adjustment how many (such as verifying whether FF EGM signal goes to pot) in fig. 17.But, will be appreciated that, before any adjustment is made to VT evidence count value, other criterions can be applied.Such as, if FF EGM signal goes to pot, then can skip this heart beating completely when not adjusting VT evidence count value.

In shown process, when FFMS falls into whole SVT district (that is, SVT confidence district 802 or SVT grey area 804), application rhythm of the heart breakpoint rule.The application of an embodiment of rhythm of the heart breakpoint rule is described in conjunction with Figure 16 hereinbefore.Compared with district's dedicated rules, rhythm of the heart breakpoint rule is the rule of higher level.Thus, if be found to be very in decision box 810 rhythm of the heart breakpoint rule, then adjust VT evidence count value at frame 834, and without the need to applying additional district's dedicated rules.In illustrated examples, in response to the evidence (that is, non-continuous tachycardia) of rhythm of the heart breakpoint, by the counter O reset of VT evidence.

When rhythm of the heart breakpoint rule is false and FFMS falls into SVT confidence district (top set of frame 810), this process advances to frame 814 to apply the special heart beating characterization rules in SVT confidence district.In an illustrated embodiment, as hereinbefore described by composition graphs 13A, apply NF abnormal heartbeats rule at frame 814.When NF abnormal heartbeats rule is fictitious time, reduce in frame 832, VT evidence count value.The disappearance of the strong evidence based on the SVT heart beating of FFMS and the VT heart beating evidence based on NF abnormal heartbeats rule proves that the relatively large decrement of VT evidence accumulation is proper.In this example, 2 are subtracted in frame 832, VT evidence count value.

If be true in frame 814NF abnormal heartbeats rule, then in specific heart beating feature, find evidence conflicting with the result of overall FFMS.Fall into SVT confidence district in response to FFMS, VT evidence still can reduce, but the decrement that compared with being fictitious time with NF abnormal heartbeats rule, reduction one is less.Such as 0.5 is reduced in frame 830, VT evidence count value.

When rhythm of the heart breakpoint rule is false (frame 810) and FFMS falls into SVT grey area (inferior division of frame 810), this process moves to frame 816 with to application area, SVT grey area dedicated rules.In one embodiment, as hereinbefore described by composition graphs 13B, VT heartbeat is applied at frame 816.When VT heartbeat is fictitious time, at frame 838, fall into SVT grey area and the disappearance of the specific heart beating feature evidence of VT heart beating in response to FFMS, VT evidence count value reduces 0.5.At frame 838, compared with when falling into VT confidence district at FFMS and there is no the evidence of abnormal heartbeats, the decrement that VT evidence accumulation reduction one is less.

When being true time at frame 816VT heartbeat, falling into the FFMS of SVT grey area and causing the increase of the VT evidence at frame 836 based on the combination of the contradiction evidence of the VT heart beating of specific heart beating feature.In the example shown, the incremental change applied at frame 836 is 0.625.

During except falling into SVT grey area at FFMS except application rhythm of the heart breakpoint rule, no matter when FFMS falls into SVT grey area and applies the strict heart rate of PVC at frame 812 and start rule (frame 804).It is also multi-region rule that the strict heart rate of PVC starts rule, and across comprising SVT grey area (frame 804) and VT grey area (frame 806) both whole grey area application.If it is true for starting rule at the strict heart rate of frame 812PVC, then at frame 840, adjust VT evidence immediately and without the need to applying any single district rule.In one embodiment, it is true for starting rule in response to the strict heart rate of PVC, and VT evidence count value is maximized.

As previously mentioned, the multi-region rule applied at frame 810 and 812 can exclusive mode write to make to only have single rule to can be very for given heart beating, or these multi-region rules hierarchical mode can be applied that VT evidence count value is excited in response to the first rule and adjust immediately.Alternatively, can comprise in flow chart 800 for verifying that before being the true VT of adjustment evidence in response to multi-region rule other multi-region rules are false additional step.For the sake of clarity, this Class Options various is not shown in fig. 17.Generally speaking, VT evidence tolerance excites adjustment in response to single rule, and does not do more than adjustment once for current heartbeat.

If it is false, then described above in application area, frame 816 pairs of SVT grey area dedicated rules for starting rule at frame 812 strict heart rate of PVC when FFMS falls into SVT grey area (top set).If fall into VT grey area and the strict heart rate of PVC starts rule is false (inferior division) at frame 812FFMS, then apply single district rule in frame 818 pairs of VT grey area.In this example, as hereinbefore described by composition graphs 13C, application FF normal heartbeat rule.If FF normal heartbeat rule is true, then, when adjusting VT evidence count value, this evidence of SVT heart beating surmounts the impact that FFMS falls into VT grey area.At frame 842, be true in response to falling into VT grey area conflicting FF normal heartbeat rule with FFMS, VT evidence reduces a relatively little decrement.

In this example, for given FFMS district, more than one district dedicated rules can be applied.For VT grey area, if be not true in frame 818FF normal heartbeat rule, then also apply as the NF normal heartbeat rule hereinbefore described by composition graphs 13C at frame 820.When finding the evidence of normal heartbeat in NF signal, for the accumulation of VT evidence, this result surmounts the impact that FFMS falls into VT grey area.At frame 844, be that true VT evidence reduces in response to NF normal heartbeat rule, instead of increase in VT grey area in response to FFMS.Will be appreciated that, the hierarchical order application that FF normal heartbeat rule (frame 818) and NF normal heartbeat rule (800) can be different, thus based on exciting as genuine first rule adjustment VT evidence.

In this example, compared with FF normal heartbeat evidence, NF normal heartbeat evidence is considered to the stronger evidence of SVT heart beating.Thus, and being truth ratio in response to FF normal heartbeat rule, is true in response to NF normal heartbeat rule, and VT evidence reduces a larger decrement.

If it is false and the single district rule (frame 818 and 820) being applied to VT grey area is all false that the strict heart rate of whole grey area PVC starts rule (frame 812), then increase in frame 846, VT evidence count value.Because FFMS is in VT grey area but not in VT confidence district, the incremental change being therefore applied to VT evidence enumerator is relatively little, such as 0.75.

When FFMS falls into VT confidence district (frame 808), add 1 immediately in frame 848, VT evidence count value.In other embodiments, for VT confidence district, can application area dedicated rules.But, in an illustrated embodiment, the very low strong evidence being considered to VT heart beating of FFMS, thus the relatively large increase ensureing the accumulation of VT evidence.

It is conceivable that, the confidence level that the application based on given rule provides, can select the incremental change/decrement of the different sizes used at each frame 830 to 840.For given rule, the confidence level being separated VT and SVT heart beating is determined by clinical assessment.

After one of frame 830 to 848 adjustment VT evidence count value, at frame 850, this VT evidence count value and detection threshold are compared.If VT evidence count value meets or exceedes detection threshold (such as, being the threshold value of 6 in one embodiment), then occur to the conversion be sure oing state 3 at frame 854.If not yet reach detection threshold in frame 850VT evidence tolerance, then detection algorithm remains in state 2, and proceeds to next heart beating to continue accumulation VT evidence at frame 852.The detection threshold being applied to VT evidence count value is different between each embodiment, and can customize for particular patient.This threshold value also will partly depend on the size of each incremental change/decrement applied in response to each heart beating characterization rules and FFMS district.

Figure 18 is the flow chart 860 for abandoning the method in process, current heartbeat being classified as go to pot signal or the signal that do not go to pot at noise/pseudomorphism.The process shown in Figure 18 of execution is to get rid of the interference of non-physiological signal to correct heart rhythm classification.Noise/pseudomorphism that method shown in flow chart 860 may correspond in being represented by Figure 13 A-C center 512,524 and 554 respectively abandons process.As mentioned above, when FF EGM signal be identified as go to pot heart beating time, can skip current heartbeat and do not use it for upgrade VT evidence enumerator (frame 556 see the frame 510 of such as Figure 13 A, the frame 526 of Figure 13 B and Figure 13 C).If NF EGM signal goes to pot, then utilize NFMS or may be false (frame 572 see the frame 514 of such as Figure 13 A, the frame 532 of Figure 13 B and Figure 13 C) from the heart beating characterization rules of the feature of NF EGM signal extraction.

Go to pot signal in order to current heartbeat is classified as thus accumulating VT evidence or judging that specific heart beating characterization rules abandons this current heartbeat as fictitious time, can check that n fragment second (n-second segment) of one or more cardiac cycle or EGM signal is to detect the existence of non-physiological signal.In other words, current heartbeat is classified as the current RRI that destroyed noise is not limited to check interested EGM signal, but inspection longer EGM signal interval can be comprised to detect the existence can disturbing the noise of correct heart rhythm classification.

Process shown in flow chart 860 can be applicable to FF EGM signal or NF EGM signal or two signals.At frame 862, sense the cardiac event used when frame 861 arranges noise corrupted analysis window.In one embodiment, the NF event sensed for arranging the one second window being applied to FF and NF EGM signal is destroyed with detection noise.N window second is configured to expand than current sensed event Zao n and terminate after NF event second.The predefine noise corrupted analysis window persistent period is set to be longer than RRI (such as, persistent period at least one second) usually, thus will there is overlap the n noise analysis second window from a heart beating to next heart beating.

At frame 866, between n window phase second, calculate the general morphology parameter of FF and/or NF EGM signal.The parameter for detection noise destruction calculated can be described as " general morphology " parameter, because these parameters can calculate in whole n signal segment second, and without the need to this analysis being limited to the morphological analysis window being used to calculate configuration mark.EGM signal base line, T ripple or other signal segment and even more than one QRS form can be comprised in n window second for measuring general morphology parameter." totally " form of EGM signal is used to detection noise/pseudomorphism and destroys, because noise artifact can occur any time in EGM signal, and is not limited to the time window corresponding with QRS waveform.Even if there is sizable overlap between n fragment second continuously, the current RRI only ending at the sensed event triggering current n noise analysis second window just can be classified as based on the noise information be included in current n fragment second and go to pot or do not go to pot.The noise falling into other previous RRI of n fragment second is sorted out not by the impact that the noise of the general morphology parameter being derived from current n fragment second is sorted out.

General morphology parameter for detecting the existence of non-physiological noise can comprise jamtosignal (NSR) (or on the contrary, signal to noise ratio), average period (MP) and the tolerance relevant to muscle noise content and be associated with the conditions associated characteristics of signals that goes between.These parameters are used to abandon the signal polluted by high-frequency noise, significant muscle noise and distinctive lead-in wire associated artifact.This noise/pseudomorphism abandons process can be included in general disclosed method in the open No.2007/0239048 of the common U.S. transferred the possession of, and this patent application by reference entirety is incorporated into this.

At frame 868, use at least current RRI, all RRI occurred during n band second or determine current HR based on the current HR estimated value of the n-th the shortest RRI in a nearest m RRI as mentioned above.If HR is greater than SVT limit value (frame 870), then the noise measuring used during frame 874 is applied in high heart rate rule.If HR is less than SVT limit value, then at frame 872, different noise measuring regular collections can be applicable to the heart rate lower than SVT limit value.

The estimated value of Current heart rate is can be depending on for the general morphology parameter that detects the signal that goes to pot and/or the threshold value that is applied to these general morphology parameters.Different noise measuring criterions can be applicable to detection noise during heart rate is higher than SVT limit value but not during heart rate is lower than SVT limit value and destroys.Usually, during HR is greater than SVT limit value, apply harsher noise corrupted criterion.Such as, higher threshold value can be applicable to general morphology parameter and destroys with detection noise.

If as frame 876 definitely any one general morphology parameter exceed noise corrupted threshold value, then at frame 880, this signal is classified as the signal that goes to pot.If these parameters all do not exceed the noise corrupted threshold value applied based on heart rate, then indicated by frame 878, this signal is not destroyed.At frame 876, can compare comprising average period, the tolerance of muscle noise and/or the noise measuring parameter of NSR disclosed in as in cited above ' 048 open application with noise measuring threshold value.Different criterions can be applicable to FF and NF signal, thus this signal is classified as go to pot signal or the signal that do not go to pot.

As hereinbefore described by composition graphs 13A to 13C, if be found to be go to pot for current RRI interval (using n fragment second) FF signal, then VT evidence count value can not be adjusted for current heartbeat.If FF signal is found to be do not go to pot, then can adjust VT evidence count value according to the result of FFMS district and any heart beating characterization rules being applied to this district.If heart beating characterization rules relates to the analysis to NF EGM signal, then when frame 880 for current heartbeat NF EGM signal be classified as go to pot time, this rule can be found to be vacation.

At frame 882, maintain the counting classifying as not destroyed n number of fragments second in the nearest fragment of specified quantity.In one embodiment, if at least half in n fragment second (such as, at least four in nearest eight n fragments second) is classified as do not go to pot, then at frame 884, rhythm of the heart mark can be treated and be set to high.This can treat rhythm of the heart mark needs for high so that allow to be transformed into from state 2 to be sure of state 3.Can treat rhythm of the heart mark is the instruction (can treat rhythm of the heart object for reliably being classified as by the rhythm of the heart when also meeting other criterions) that analyzed EGM signal is considered to the signal of sufficiently clear.Such as, in one embodiment, if NF and FF HR estimated value is both more than SVT limit value and NF can treat rhythm of the heart mark and FF can treat rhythm of the heart mark and be set to high, then the conversion be sure oing state 3 can be occurred to.Even if VT evidence count value is not crossing with detection threshold, also can occur based on NF and FF HR this conversion more than SVT limit value.

If (or another percentage ratio) over half recently in n fragment second is classified as go to pot, then at frame 886, the treated rhythm of the heart mark of corresponding EGM signal is set to low.In the case, be transformed into from correlation behavior 2 other criterions be sure oing state 3 even if meet, due to the destruction of EGM signal, also State Transferring can not occur.Such as, if to proceed to, NF and FF HR estimated value be sure of that but state 3 NF or FF EGM signal can be treated one of rhythm of the heart mark and is set to low more than SVT limit value, then will not there is State Transferring.Before detection algorithm will proceed to and be sure of state 3, at least one or two in FF and NF EGM signal, be classified as based on respective signal and do not go to pot, and rhythm of the heart mark can be treated and must be set to high.

In certain embodiments, based on the state can treating rhythm of the heart mark, can switch between NF and FF EGM signal the dependence of " mainly " EGM signal.Such as, if NF EGM signal is found to be and does not go to pot and NF can treat rhythm of the heart mark is set to low, then FF EGM signal becomes for sensing cardiac event, arranging morphological analysis and calculating the main signal of configuration coupling mark.Do not go to pot when NF EGM signal is found to be again (that is, NF can treat rhythm of the heart mark and be set to high) time, again recover it as the effect of main sensing signal sensing cardiac event and arrange morphological analysis window.

If FF EGM signal is found to be and suffers noise corrupted and FF can treat rhythm of the heart mark is set to low, then NF signal can be used for calculating configuration coupling mark, until FF can treat rhythm of the heart mark be again set to high.Equally, can use and be only derived from it and can treat heart beating feature that rhythm of the heart traffic sign placement is high EGM signal to apply specific heart beating characteristic rule.Such as, can replace FF signal characteristic in the specific heart beating characterization rules usually depending on NF signal characteristic, vice versa.

In one embodiment, if FF EGM signal goes to pot, then replace as the object for the accumulation of VT evidence hereinbefore described by composition graphs 13A-13C and skip this heart beating, can NFMS be calculated.If NFMS is in confidence district, then in response to NFMS district adjustment VT evidence count value.If NFMS is in grey area, does not then adjust VT evidence count value and skip this heart beating.Compared with the incremental change applied in response to FFMS confidence district (when FF EGM does not go to pot) or decrement, in response to NFMS confidence district (when FF EGM goes to pot), a relatively little incremental change or decrement be can be to the adjustment amount of VT evidence count value.In illustrated examples, when NFMS in VT confidence district and FF EGM signal be found to be go to pot time, VT evidence count value can increase by 0.75 instead of 1.When NFMS in SVT confidence district and FF EGM signal goes to pot time, VT evidence count value can reduce 1.5 instead of 2.

Figure 19 be for calculate EGM signal is classified as go to pot signal or do not go to pot signal time the flow chart 900 of a kind of method of the tolerance of lead-in wire pseudomorphism that uses.Abandon a part for process as the noise/pseudomorphism shown in Figure 18, can to respective signal is classified as go to pot signal or do not go to pot signal time FF or the NF EGM signal that uses in any one or two flowchart 900 shown in process.At frame 902, sense the cardiac event (R ripple) about NF EGM signal, for arranging n noise analysis second window at frame 904.

At frame 906, all zero crossings (zero crossing) during n fragment second of location.The sampled point that mark and each zero crossing are close to.The amplitude of the neighbouring sample point after the amplitude of the neighbouring sample point before zero crossing and same zero crossing is compared.At frame 908, the sampled point with minimum absolute amplitude adjacent with each zero crossing is set to zero amplitude, thus is null value by this grappling, for the pulse dividing positive and negative trend in n noise analysis second window.

Then, rectification is carried out in frame 910 pairs of n signal segment second.Determine the peak swing through rectified signal, and at frame 912, calculate the maximum predetermined percentage through rectified signal amplitude or part (such as, the maximum half through rectified signal amplitude) as pulse amplitude threshold value.At frame 914, its amplitude is greater than to all counting through rectified signal sampled point of the maximum half through rectified signal amplitude (or another percentage ratio).At frame 916, this counting is stored as noise artifact tolerance, tolerance 1.In one embodiment, noise artifact tolerance is used as the measurement to lead-in wire associated artifact.Can be used as the maximum percentage ratio through rectified signal amplitude or the part of pulse amplitude threshold value based on HR definition, this percentage ratio or part can be used in any method described herein and estimate.When HR is greater than SVT limit value, apply harsher noise measuring criterion, with calculating noise Artifact Metric 1 by selecting the maximum major part through rectified signal amplitude as pulse amplitude threshold value.

At frame 918, identify the individual pulse that having in rectified signal exceedes the sampled point of the maximum quantity of the maximum half through rectified signal amplitude (or another percentage ratio).In order to identify this individual pulse, the sampled point exceeding the maximum half through rectified signal amplitude in each independent pulse during n fragment second is counted.Then, the counting of each independent pulse and the counting of each other pulse are compared.The pulse with the sampled point of the maximum quantity exceeding the maximum half through rectified signal amplitude is identified as the pulse with maximum pulse.At frame 920, store the quantity of the point of the maximum half through rectified signal exceeded in the pulse of maximum pulse as the second lead-in wire Artifact Metric.

At frame 922, lead-in wire Artifact Metric is compared with the signal threshold value that goes to pot of the function being defined as the second lead-in wire Artifact Metric (" pseudomorphism 2 ").Such as, lead-in wire Artifact Metric 1 can be compared with the threshold value being defined as W* (measuring 2-x).Item as used herein " W " and " x " need not equal herein by other member of equations or relevant to it of same letter mark.Weighter factor " W " and/or item " x " can be set based on heart rate (such as, as described in conjunction with Figure 18, based on current HR estimated value whether more than SVT limit value).When HR is greater than SVT limit value, noise threshold is set to high value, to create harsher noise measuring criterion by increasing " W " and/or " x ".

In one embodiment, if the sum exceeding the sampled point of the half of peak swing in rectified signal is greater than 10 times that exceed the quantity of the sampled point of the half of peak swing in individual pulse, then at frame 926, for current heartbeat, this signal is classified as the signal that goes to pot.If do not exceed at frame 922 Artifact Metric that goes between the signal threshold value that goes to pot, then for current heartbeat, before EGM signal being classified as go to pot signal or the signal that do not go to pot, other general morphology tolerance can be analyzed at frame 924.As previously mentioned, other general morphology tolerance can comprise NSR, average period and the tolerance relevant to muscle noise.

Figure 20 is the flow chart 700 of the method for changing between detected state.Discuss the method for being transformed into correlation behavior 2 from state 1 (uncorrelated state) hereinbefore.Once get the hang of 2 at frame 702, just on the basis of heart beating one by one, accumulate VT evidence.If as frame 704 institute definitely VT evidence count value do not reach detection threshold, then start at frame 706 condition for converting back state 1 of checking.Generally speaking, if HR reduces and VT evidence drops to and stops below threshold value, then detection algorithm convertible time uncorrelated state 1.During switching criterion can be depending on and is LV operator scheme state 1 from state 1 to the conversion of state 2 or occur during its HV operator scheme.Flow chart 700 provides a kind of illustrative method that can be used for controlling to convert back state 1 from state 2.

At frame 706, current RRI and detection interval lower limit are compared.If RRI is than limit for length under detection interval, then (switchback) count value at frame 716, is turned to increase.Turn to enumerator to be used to follow the tracks of long RRI (that is, being greater than detection interval lower limit) all the time and convert back state 1 for control.

If RRI is not more than detection interval lower limit and got the hang of for 2 (as determined at frame 712) from state 1 during LV operator scheme, then at frame 714, current RRI and the expection RRI scope stored are compared.As described earlier, expection RRI scope when being transformed into state 2 from state 1 is frozen in its currency, and is stored to use when controlling to convert back state 1.If at the current RRI of frame 714 not within the scope of expection RRI, then at frame 708, turn to count value to reduce.

In addition, at frame 710, slow heartbeat counting value reduces.Slow Heart rate meters is and turns to enumerator independently enumerator, and for counting " slowly " heart beating.Slow Heart rate meters is used to follow the tracks of slow heart beating all the time during state 2 and 3 and directly converts back state 1 for controlling from state 3, as will be described further below.In response to shorter than Monitoring lower-cut and not within the scope of the expection RRI of normal cardiac rhythm RRI, turn to count value and slow heartbeat counting value can reduce 1,2 or decrement selected by another.

To be current cardiac frequency remain the relevant rhythm of the heart to and RRI still beyond normal, expected RRI scope shorter than Monitoring lower-cut and ensure the evidence that do not have state to change.After the slow heartbeat counting value of reduction, by being back to frame 702, detection algorithm remains in state 2.This rhythm of the heart is still considered to the relevant rhythm of the heart.

Be back to frame 714, if current RRI is within the scope of the expection RRI of stored normal cardiac rhythm, then at frame 716, turn to count value to increase.At frame 718, in response to RRI (frame 706) longer than Monitoring lower-cut when occurring to the conversion of state 2 during the LV operator scheme of the state 1 or RRI within the scope of expection RRI, slow heartbeat counting value also increases.

At frame 720, count value and switching threshold will be turned to compare.If turn to count value to reach switching threshold, then VT evidence and termination threshold value are compared (frame 722).Stop threshold value and be defined as the value less than detection threshold, and be used to determine when that accumulated VT evidence no longer meets the level of the relevant rhythm of the heart of instruction.If VT evidence count value is lower than termination threshold value, then VT outbreak may no longer occur forever, or non-continuous VT outbreak can occur and spontaneously stop.

If any one in current NF or FF HR is less than SVT detect limit value (frame 724), then at frame 726, the low accumulation turning to counting, VT to count of slowly heart beating all the time and the combined evidence of the HR estimated value lower than SVT detection limit value cause converting back uncorrelated state 1.If do not meet any one in these switching criterions, then detection algorithm remains on (frame 730) in state 2.

Be back to frame 712, if do not got the hang of for 2 (that is, from HV mode entering status 2) from the LV operator scheme state 1, then analyze RRI variability at frame 732.In order to get the hang of 2 from HV operator scheme, detect the unexpected reduction being attended by the variability that HR increases.If RRI variability increases again, then current cardiac frequency may be no longer the relevant rhythm of the heart.When from the HV mode entering status 2 of state 1, expection RRI scope can not be stored.High variability due to RRI causes wide in range desired extent, no longer can upgrade expection RRI scope during HV pattern.Thus, when from the HV mode entering status 2 of state 1, check that RRI variability is to control to convert back state 1 from state 2.

At frame 732, by the meansigma methods of the one or more nearest RRI difference between two successive heartbeats and nearest RRI being compared, the increase of RRI variability can be detected.If average or other tolerance of RRI difference are greater than the average predetermined percentage (such as, about 20%) of RRI recently, then detect the increase of RRI variability at frame 732.

If RRI variability increases, then at frame 734, current RRI and tachycardia are expected that RRI scope compares.As mentioned above, getting the hang of after 2, initiate tachycardia expection RRI scope, and on the basis of heart beating one by one, use current RRI and previous RRMEAN and RRMAD to upgrade this tachycardia expection RRI scope.If current RRI is than tachycardia expection RRI scope long (frame 734), then in response to the RRI variability increased and RRI (slower than this tachycardia desired extent) beyond tachycardia desired extent, detection algorithm is transformed into uncorrelated state 1 (frame 726).

If at frame 732 and 734 when from the criterion do not met during HV mode entering status 2 for switching back state 1, then at frame 708, turn to count value to reduce.At frame 710, slow heartbeat counting value reduces.Equal in response to current RRI or be shorter than detection interval lower limit, turning to and reduce with slow heartbeat counting value, this instruction current cardiac frequency keeps the relevant rhythm of the heart.

Generally speaking, the RRI variability of, increase among the expection RRIs of normal cardiac rhythm or than the expection RRI scope of this normal cardiac rhythm long, than tachycardia the desired extent long evidence of RRI, the VT evidence of low accumulation and/or state 2 during longer than Monitoring lower-cut all the time or its combination in any can be used for control and are transformed into uncorrelated state 1 from correlation behavior 2.

Refer again to frame 704, if VT evidence reach detection threshold (such as, if VT evidence reach 6 or another predefine detection threshold), then occur to the conversion (frame 740) be sure oing state 3.In addition, when FF and NF HR estimated value both more than SVT heart rate limit value and FF and NF EGM signal both based on noise/pseudomorphism abandon analysis be classified as can treat time, no matter all can occur to the conversion of state 3 why in frame 740 VT evidence count value.

During state 3, perform therapeutic choice process to determine should transmit what treatment (if any) in response to the VT detected at frame 742.Continue in the mode identical with state 2 in the accumulation of frame 744, VT evidence.Depend on treatment decision process and selected treatment, get the hang of 3 and treatment transmission start between can life period postpone.At this time durations, the accumulation of VT evidence continues, thus the spontaneous termination of VT that detects of identifiable design or the pause of the rhythm of the heart.

At frame 746, VT evidence count value and termination threshold value are compared, this termination threshold value is defined as the value less than detection threshold.In one embodiment, detection threshold is that 6 to stop threshold value be 2.If VT evidence count value reaches termination threshold value and NF or FF HR drops to below SVT heart rate limit value (frame 748), then detection algorithm can convert back state 2 from state 3.Stop the threshold value that threshold value is not interpreted as stopping for detecting episode of tachycardia, but be interpreted as the threshold value for detecting the needs being transformed into lower detected state.

Before being transformed into state 2 from state 3, adjustable and the slow heartbeat counting value of inspection are to determine whether to ensure directly to convert back state 1 from state 3.If current RRI is than limit for length under detection interval (frame 752), then at frame 756, slow heartbeat counting value increases.At frame 758, slow heartbeat counting and state change threshold are compared.If slowly heartbeat counting value reaches state 1 change threshold, then there is the conversion (frame 726) of directly arriving state 1.The combined information of low VT evidence (frame 746), the current RRI (frame 752) than limit for length under detection interval and the lasting slow heart beating (frame 758) based on slow heartbeat counting ensures the conversion of uncorrelated state 1.

If not yet reach state 1 change threshold at frame 758, then this rhythm of the heart is still considered to the relevant rhythm of the heart.At frame 702, detection algorithm converts back state 2.

On the other hand, if shorter or equal this Monitoring lower-cut than Monitoring lower-cut at the current RRI of frame 752, then at frame 754, slow heartbeat counting value reduces.But based on low VT evidence and NF or FF HR still shorter than detection interval lower limit than the low current RRI of SVT heart rate limit value, make the conversion getting back to correlation behavior 2.Before turning back to state 1 or again advancing to state 3, perform additional monitoring.

Made after treatment judges at frame 742, if the VT evidence accumulated remains to be greater than stop threshold value (be negative decision at frame 746), then detection algorithm proceeds to state 4 (frame 750).At frame 766, in state 4, transmit treatment as scheduled.After treatment transmits, at frame 768, by the counter O reset of VT evidence.

When transmitted treatment is impact treatment, FF EGM signal can be unreliable within the several seconds.A Time Intervals can be set at the unreliable timer of frame 770, FF, thus allow to impact the minimizing before FF EGM signal detects for tachycardia again of after-polarization (polarization) pseudomorphism.Similarly, whenever transmit shock pulse or any other treatment (such as bringing out the T wave impact of VF) for diagnoses and treatment object, the unreliable timer of FF can be set.The unreliable timer of FF can be set to the interval of such as about 2 to 5 seconds.Similarly, if transmitting in treatment (such as pacing therapy) electrode used for sensing NF signal, then the unreliable timer of NF can be configured to allow polarization pseudomorphism and/or any NF EGM metamorphosis to dissipate before NF signal detects for tachycardia again.

The conversion (frame 702) of state 2 is got back to immediately after treatment transmits.Detection algorithm is transformed into state 2 to allow to monitor the continuation of the rhythm of the heart.Treatment transmit after the back to back rhythm of the heart be still considered to the relevant rhythm of the heart because this treatment not yet success or VT outbreak may heavily come.Thus, when be back to state 1 (or again detect, recurrence or worsen VT state 3) before transmit treatment after, carry out the monitoring to the rhythm of the heart in state 2.

Figure 21 is the flow chart 950 of operator scheme after the treatment performed after reentering state 2 after transmitting treatment.In certain embodiments, after transmission treatment, differently operation detection algorithm can be operated with state 2 before treatment and state 3.Can be applied in during operator scheme after treatment in state 2 after transmitting treatment for again detect VT outbreak unique criterion and for detecting the criterion that VT outbreak stops.

At frame 951, in state 4, transmit treatment, and the conversion making the state of getting back to 2 treats rear operator scheme (frame 952) to enter.Getting the hang of after 4, no longer upgrade tachycardia desired extent.Being stored in the last tachycardia desired extent calculated before treatment transmits at frame 954, using, as further described hereinafter when again detecting tachycardic criterion for being used in application.In certain embodiments, at frame 954, be only stored in the average RRI used when tachycardia desired extent is set, as the threshold value RRI used when again detecting tachycardia.

As described in conjunction with Figure 20, when transmitted treatment is impact treatment, the unreliable timer of FF EGM can be set, thus at least not use FF EGM signal at first during operator scheme after the treatment.On the contrary, NF EGM signal is used to sense cardiac event, thus determines RRI, arranges morphological analysis window and determine configuration score and corresponding form mark district, to adjust the VT evidence count value after treatment.If the treatment transmitted is pacing therapy (that is, anti-tachyarrhythmia pacemaker (ATP) treatment), then FF EGM signal can continue on for determining configuration mark.Thus, after the treatment during pattern, depend on transmit treatment type, different signal processing methods can be used.

At frame 956, on the basis of heart beating one by one, be based upon the analysis window that current NF sensed event is arranged, determine configuration mark according to FF EGM signal (after ATP) or NF EGM signal (after impacting).As described earlier, according to form Matching band, each form mark is sorted out.At frame 958, upgrade VT evidence count value based on form mark district.

In certain embodiments, when transmitted treatment is impact treatment, do not apply specific heart beating characterization rules during pattern after the treatment.Specifically, at least until the unreliable timer expires of FF or until detecting that termination reenters normally " before treatment " operator scheme, just application depends on any rule of the feature determined according to FF EGM after the blow.After the blow, only based on NFMS district adjustment VT evidence count value.

Alternatively, the heart beating characterization rules of limited quantity is applied after the blow.In one embodiment, like the FF normal heartbeat Regularia hereinbefore described by composition graphs 13C, normal heartbeat rule just impacted district's dedicated rules of rear application before adjustment VT evidence count value.When configuration mark falls into VT grey area, only use specific NF heart beating feature or after the unreliable timer expires of FF, only use FF heart beating feature to apply normal heartbeat rule.If normal heartbeat rule is true, then in response to configuration mark in VT grey area, VT evidence count value can reduce but not increase (or being the less incremental change of increase one compared with fictitious time with normal heartbeat rule).

Additionally or alternatively, after the treatment during operator scheme, can across the application of one or more form districts and rule like previously described rhythm of the heart breakpoint Regularia.Such as, if long NF RRI detected and the form mark terminating this long RRI falls into SVT confidence district or is greater than another rhythm of the heart breaking point detection threshold value, then rhythm of the heart breakpoint can be detected.If rhythm of the heart breakpoint detected, then remove VT evidence enumerator at frame 958.Similarly, if two continuous print form marks are in SVT confidence district, then VT evidence enumerator can be removed at frame 958.

After adjustment VT evidence count value, the NF HR after treatment estimated by frame 960.As mentioned above, NF HR can be estimated based on the n-th the shortest RRI in the nearest RRI of the predetermined quantity occurred after treatment transmission.From the n-th available heart beating, the NF HR estimated value after treatment the earliest can be determined.Such as, if the 9th RRI the shortest in nearest 12 RRI will be used as HR estimated value, then the shortest RRI identified after 9 RRI will after the treatment be first NF HR estimated value.After the n-th treatment before RRI, the renewal to VT evidence count value can be started on the basis of heart beating one by one.

Apply during operator scheme after the treatment for again to detect and/or the criterion of termination detection depends on NF HR estimated value.Thus, at frame 962, NF HR estimated value and SVT heart rate limit value are compared.If NF HR is faster and NF signal does not go to pot (frame 964) than SVT limit value, then at frame 966, again detect episode of tachycardia in response to high HR.At frame 968, occur to the conversion be sure oing state 3.If pattern enter treatment after ATP treatment after, then FF EGM signal can be used for checking NFHR estimated value as mentioned above.If FF EGM signal after the treatment during pattern (such as, after ATP) use, then at frame 964, before being transformed into state 3, FF signal also can be verified as and not go to pot.

At frame 968, during state 3, detection algorithm, according to operator scheme after treatment, the basis of heart beating is one by one continued upgrade VT evidence count value, makes therapeutic choice simultaneously.During state 3, also upgrade NFHR estimated value, thus detect the termination of episode of tachycardia during state 3.Thus, during operating after the treatment of state 3, if make treatment judge and treatment transmit ready before meet stop criterion (such as, continue RRI than limit for length under detection interval), then can there is the conversion directly getting back to state 1.

Refer again to frame 962, if but as frame 974 definitely NF HR estimated value be less than SVT limit value and be greater than sudden change limit value, then at frame 976, VT evidence count value and detection threshold are compared.If VT evidence count value is greater than detection threshold (frame 976) and NF signal (and the FF signal after ATP) does not go to pot (frame 964), then again detect episode of tachycardia at frame 966.

If NF HR is greater than sudden change limit value, then the termination of episode of tachycardia cannot be detected.Detection algorithm will remain in correlation behavior 2 (or state 3).Thus, when NF HR is greater than SVT limit value (frame 962) or be less than SVT limit value but be greater than sudden change limit value (frame 974), applies detection criteria again at frame 976 and/or 964, but do not apply termination detection criterion.If less than the new detection criteria of heavy sensation in the foot (negative branch of frame 976 and 964), this process is back to frame 956 to sense next cardiac event and the NF HR determining next configuration mark district, upgrade VT evidence count value and upgrade through estimating.

In order to meet detection criteria again when being greater than SVT limit value at NF HR or being greater than sudden change limit value, NF signal (and adjustment VT evidence rely on FF signal) should be not destroyed.At frame 964, according to the state that such as " can treat the rhythm of the heart " described in conjunction with Figure 19 indicates, NF and FF signal can be confirmed as go to pot signal or the signal that do not go to pot.

If NF HR estimated value is less than sudden change limit value (negative decision at frame 974) but is greater than tachycardia Monitoring lower-cut (frame 980), then application termination detection criterion and again detection criteria.At frame 982, termination detection criterion can comprise comparing of current RRI and detection interval lower limit.Termination detection also may need detection algorithm in state 2 or state 3 operation amount at least n second (such as, about 3 seconds) to avoid State Transferring frequently.In response at frame 982, termination detection criterion is met for current RRI, stop count value at frame 988 and increase.

Figure 22 is the flow chart 1001 for detecting a kind of method that the rear VT for the treatment of stops according to an embodiment.Method shown in Figure 22 may correspond to the operation for applying termination detection criterion performed in the frame 982 at Figure 21.Be less than sudden change limit value (frame 1002) when NF HR estimated value but when being greater than Monitoring lower-cut, at frame 1004, apply stop criterion by current RRI and normal, expected RRI scope being compared.Before VT outbreak detects, normal, expected RRI scope is the final expection RRI scope of storage after never correlation behavior 1 is transformed into correlation behavior 2.If at the current RRI of frame 1004 within the scope of normal, expected (or longer than this normal, expected scope), then at frame 1008, stop heartbeat counting value and increase.

If current RRI and detection interval lower limit not within the scope of normal RRI or longer than this normal RRI scope, then at frame 1006, alternatively compare by current RRI.If exit status 1 during HV operator scheme, then can not store normal, expected RRI scope.In the case, the threshold value of detection interval lower limit can be used at frame 1006, instead of at frame 1004, itself and expection RRI scope be compared.

If current RRI is greater than at least one in the expection RRI scope or detection interval lower limit finally stored, then at frame 1008, stops heartbeat counting value and increase.If do not meet the condition of testing at frame 1004 and 1006, then at frame 1010, stop heartbeat counting value and reduce.

Within the scope of the expection RRI that current RRI in the end stores or the expection RRI scope of storage more last than this is long or than detection interval under limit for length time, stop heartbeat counting value and can increase one or another incremental change.Whenever current RRI than normal, expected RRI scope or detection interval lower limit in short-term, stop heartbeat counting value and can reduce two or another decrement.

Increase if stop count value at frame 1008, and as frame 1014 definitely in the current detection algorithm state (state 2 or state 3) during operation at least n to expire second, then the additional stop criterion of application.If detection algorithm not yet operates at least n second (such as, about 3 seconds) in current state, then termination can not be detected.At frame 1014, the Time Intervals that can be applicable in current state needed for operation prevents State Transferring frequently.When not meeting stop criterion, at frame 1012, this algorithm can continue to apply detection criteria again.Tout court with reference to Figure 21, when not meeting stop criterion at frame 982, this process proceeds to the frame 984 of Figure 21 to apply detection criteria again.

At Figure 22, if the required time interval in frame 1014 current state expires, then at frame 1016, VT evidence count value and termination threshold value are compared.If VT evidence count value has dropped to below the selected threshold value of VT evidence, then at frame 1018, termination heartbeat counting value and termination heart beating threshold value are compared.If stop heartbeat counting value reach threshold count, then termination detected at frame 1020.All the time grow within the scope of expection RRI or than this expection RRI scope in response to low VT evidence and RRI or than limit for length under detection interval, termination detected.

If VT evidence count value or termination heartbeat counting value do not meet corresponding termination detection threshold value, then at frame 1024 and 1026, additional stop criterion can be applied.At frame 1024, the HR that current NF HR estimated value is corresponding with the tachycardia desired extent finally stored compares.In one embodiment, if at frame 1024, the n-th the shortest RRI in the RRI after nearest m treatment grows to few 50ms than tachycardia desired extent, then at frame 1026, if the pattern of variable RRI detected, then still can termination be detected.By length, rule (regular) or stable RRI characterize serious VT to heavens.Thus, termination detection criterion can comprise and detect unstable RRI as the relevant criterion of the instruction being back to the non-ill rhythm of the heart.

In one embodiment, the Cumulate Sum of the continuous RRI difference of RRI after the nearest treatment of selected quantity is compared with the percentage ratio of the meansigma methods of identical RRI.Such as, if the Cumulate Sum of RRI difference be greater than average RRI at least about 10%, then this RRI is considered to unstable, namely indicates the untreatable rhythm of the heart.

This RRI unstability combined with the HR being less than tachycardia desired extent is considered to VT by the evidence of the treatment being transmitted successful termination.Termination is detected at frame 1020.If do not meet the criterion for detecting VT termination applied at frame 1024 and 1026, then at frame 1028, this algorithm application detection criteria again.

With reference to Figure 21, when meeting termination detection criterion at frame 982, termination detected at frame 992, and occur to the conversion of uncorrelated state 1 at frame 994.If do not meet termination detection criterion at frame 982, then detection algorithm proceeds to frame 984 to apply detection criteria again.

Figure 23 is the flow chart 1100 being used for again detecting during operator scheme after the treatment a kind of method of VT when NF HR is greater than tachycardia Monitoring lower-cut but is less than sudden change limit value (frame 1102).At the frame 984 of Figure 21, the method for again detecting VT can be applied.

At frame 1104, VT evidence count value and detection threshold are compared.If not yet reach detection threshold, then do not carry out the detection again to current heartbeat.At frame 1108, this algorithm proceeds to next heart beating.If but VT evidence count value is more than or equal to detection threshold NF EGM signal (or the FF EGM signal after ATP), go to pot (negative branch of frame 1106), then for current heartbeat, and the discontented new detection criteria of heavy sensation in the foot.The destruction of NFEGM signal can be identified based on the state that such as " can treat the rhythm of the heart " described in conjunction with Figure 18 indicates.Tout court with reference to Figure 21, by being back to frame 956, this algorithm proceeds to next heart beating, thus determines next NFMS and NF HR estimated value, again detects and/or stop criterion to apply as required according to next NF HR estimated value.

In fig 23, when having reached detection threshold in frame 1104VT evidence count value and not gone to pot at frame 1106NFEGM signal, at frame 1110, current RRI and the expection RRI scope finally stored are compared and/or itself and detection interval lower limit compared at frame 1112.When during exit status 1, storing expection RRI scope to use when frame 1110 compares during LV operator scheme.If exit status 1 during HV pattern, then detection interval lower limit is used as frame 1112 and compares by threshold value for again detecting.If current RRI unlike normal, expected RRI scope or the interval corresponding with tachycardia Monitoring lower-cut short, be then not in relation to current heartbeat and again detect VT.At frame 1108, detection algorithm proceeds to next heart beating.

If current RRI is shorter and/or shorter than tachycardia detection interval lower limit than normal, expected scope, then calculate the instable coefficient of RRI at frame 1114.The instable coefficient of RRI is calculated as the tolerance of the difference between the rear RRI of formerly treatment of current RRI and selected quantity.What calculate as described in conjunction with Figure 22 does not rely on the trend of HR for the RRI unstability tolerance detecting termination based on continuous print RRI difference.Use discontinuous RRI difference to calculate the instable coefficient of RRI, and the trend of this coefficient to heart rate is responsive.

In one embodiment, the instable coefficient of RRI be calculated as each after current RRI and " a-1 " individual treatment recently in RRI divided by the meansigma methods of " a " individual RRI business between the Cumulate Sum of absolute difference.At frame 1116, this coefficient instable for RRI and threshold value are compared.In one embodiment, the threshold value being applied to the instable coefficient of RRI is about 0.6.If be greater than detection threshold again at the instable coefficient of frame 1116RRI, then do not carry out the detection again about current heartbeat, and at frame 1108, this algorithm proceeds to next heart beating.

If definitely RRI difference sum is less than threshold value with the ratio of RRI meansigma methods, then RRI is considered to high stability and is the instruction of possible serious VT.At frame 1120, keep being greater than detection threshold in response to VT evidence, NF EGM signal does not go to pot, current RRI is than RRI desired extent or detection interval lower limit is short and high stability RRI, again detects VT.In certain embodiments, in order to again detect VT at frame 1120, must meet at frame 1118 additional requirement that NF HR falls into the last tachycardia desired extent stored.

Referring again to Figure 21, and NF signal does not go to pot (frame 964) if meet again detection criteria at frame 984, then again detect VT, and proceed at frame 968 conversion be sure oing state 3.

For selecting during pattern after the treatment again to detect/the final HR condition of stop criterion is that NF HR is less than tachycardia verification and measurement ratio lower limit (negative branch of frame 980).In the case, at frame 982, current RRI and detection interval lower limit are compared.If current RRI is not less than detection interval lower limit, then at frame 984, stop heartbeat counting value and can reduce (or remaining on zero count).By returning frame 956, this process proceeds to next heart beating.

Under being greater than detection interval at the current RRI of frame 982 in limited time, at frame 988, for current heartbeat, stop heartbeat counting value and increase.Then, at frame 990, compare stopping heartbeat counting value with the threshold value being used for detecting termination.Stop heart beating threshold value if not yet reached, then for current heartbeat, do not detect termination.By returning frame 956, this process proceeds to next heart beating.

If stop heartbeat counting value at frame 990 reached threshold value for detecting termination, then detect at frame 992 and stop.Pass through transmitted treatment and successfully terminate VT.The conversion of state 1 is got back at frame 994.Be applied to and stop the termination threshold value of Heart rate meters and can be set to about 5 or another selected numeral, thus when under each continuous RRI is than detection interval during limit for length, termination can detected lacking to the heart beating of number of thresholds after the treatment.In certain embodiments, for again detecting and/or the object of termination detection, the RRI (such as, RRI after 2 to 3 treatments) of the fixed qty after transmitting immediately preceding treatment can be ignored.

When entering pattern after treatment after impacting transmission, various method can be used control the use when recovering FFEGM signal.Remember, when transmitting impact, after treatment, operator scheme only depends on NF EGM signal.After transmission impact treatment, before FF EGM signal can not return baseline, treatment, form reaches several seconds or even more than one minute or one minute.After detection stops, automatically can be back to FF and the NF signal processing adopting FF and NF signal to carry out the combination of controlled state conversion.

In other embodiments, recover FF EGM signal processing to stop/again detect haveing nothing to do with whether having there occurs.In certain embodiments, the time of definable fixed amount ignores FF EGM signal.In the case, NF EGM signal is used after the treatment during pattern, until timer expires.After timer expires, recover FF and the NF EGM signal processing of combination.Detection again detected or stopping and before or after corresponding state conversion, timer expires can occur.Therefore, FF and the NF EGM signal processing of combination recovery and not necessarily depends on detection algorithm state.In another embodiment, when detect termination or timer expires (whichsoever first occur or both may all need) time, FF EGM signal processing can be recovered.

Alternatively, perform with control, when two EGM signal processing is recovered to the further analysis of FF EGM signal at frame 995.Criterion can be applicable to FF EGM signal to determine when to reuse FF EGM signal.Such as, the recovery of the FF EGM signal analysis except NF EGM signal analysis may need the heart beating of specific quantity to have FFMS as fallen into SVT confidence district definitely at frame 995.Other criterions can be used for determining FF EGM signal be back to impact before baseline form with control from after impacting only NF signal processing to the conversion of dual signal FF and NF EGM signal processing.

In one embodiment; after the unreliable timer expires of FF after treatment transmits; FFMS can be determined on the basis of heart beating one by one, thus the quantity (such as, falling into the FFMS in SVT confidence district or whole SVT district) of the heart beating had normally or close to normal morphology is counted.When FF EGM signal represent constantly normal or close to normal morphology (this can use the enumerator increasing in response to FFMS or reduce to follow the tracks of) time, the recovery of FF signal processing model can be there is, no matter current detection algorithm state.

But stop thus cause converting back uncorrelated state 1 FFEGM signal at frame 994 being not yet back to baseline form (frame 995) if detected at frame 990, then detection algorithm can only use NF EGM signal to continue to operate (frame 996).NF EGM signal is only had to be used to the sudden change of detected state 1.Once FF EGM signal has been back to baseline form, operator scheme that detection algorithm just returns " before treatment ", this has depended on NF and FF EGM signal processing (frame 998) and has carried out controlled state conversion.

Should be appreciated that as shown in figure 21, recovering FF EGM signal processing may not need termination detection.If FF EGM form is found to be back to baseline form and detection algorithm remains in state 2 or state 3, even if then before termination being detected, FF EGM signal also can be used except NF EGM signal to start detection algorithm to apply heart beating characterization rules, to upgrade VT evidence count value and controlled state conversion.

Generally speaking, before the treatment in pattern, carry out FF and NF EGM signal processing, and detection algorithm criterion is applied to FF and NF signal to control each State Transferring.After the treatment in pattern, use the criterion for detecting termination (and from state 2 or state 3 to the conversion of state 1) and the criterion for again detecting (from state 2 to the conversion of state 3), these criterions differently can define with the criterion changed for controlled state before the treatment.In addition, after the treatment in pattern, the signal processing method used will depend on the type of transmitted treatment.After ATP, FF and NF signal can apply stop and again detection criteria time use.But, after the blow, only use NF EGM signal, until meet the criterion for recovering to use FF EGM.

Once detect and stop and recovered FF EGM signal processing, detection algorithm just can be described as again and operates in pattern before the treatment.If termination do not detected after the treatment, then detection algorithm can remain on after treatment in pattern, thus may again detect and one or many transmit another treatment, until meet stop criterion.When the unreliable timer expires of FF, meet form criterion or meet condition needed for other time, any time after the treatment during pattern, the recovery of FF EGM signal processing can be there is.But if the criterion stopping not meeting for recovering FF EGM signal detected, then only have NF EGM signal processing to continue, but detection algorithm reuse various " before treatment " criterion and stop and detection criteria again after non-treatment for the rule of controlled state conversion.Once termination be detected and recover FF EGM signal processing, detection algorithm just returns " before treatment " operator scheme completely.

In each flow chart presented in this article and describe, multiple criterion is described to be applied to controlling response sometimes.Response can comprise adjustment count value, transit to different detection algorithm states, in algorithm state change detection algorithmic model, mark is set, arrange intervalometer or another respond.Should be appreciated that, when multiple criterion be described to be for special-purpose is provided and demand fulfillment time, each embodiment of method described herein can independently (single criterion) or with the criterion criterion described by any Combination application, to reach the object controlling specific response.Further, in some cases, the particular order of the Applied criteria that the order applying multiple criterion can describe from flow chart presented in this article makes change criterion.It is contemplated that, when describing criterion for controlling multiple criterion of specific response, any subset of those criterions or combination can be applied before providing response.

Thus, present for detecting tachycardic method and apparatus with reference to specific embodiment in the above description.Should be appreciated that and can make various amendment to referenced embodiment and the scope of the present disclosure do not deviated from as set forth in the dependent claims.

Claims

1., for distinguishing an armarium for cardiac event, comprising:

Multiple electrode, sense the almost whole persistent period sterically defined heart signal across predetermined sensing window, the heart signal sensed comprises two-way amount heart signal; And

Processor, described processor is configured to:

Determine the configuration coupling mark corresponding to sensed heart signal and known template; Described configuration coupling mark is compared with multiple form Matching band; Select across the heart beating feature in multiple heart beating features of the whole persistent period being less than described sensing window, described heart beating feature based on described configuration coupling mark fall into described multiple form Matching band which and choose from multiple heart beating feature; Accumulation cardiac event evidence is carried out in response to described configuration coupling mark and determined heart beating feature; And distinguish cardiac event in response to accumulated cardiac event evidence.

2. equipment as claimed in claim 1, it is characterized in that, described multiple electrode comprises:

Along the first electrode pair of the first sensing vector sensing heart signal; And

Along and described first sense vector different second and senses vectorial the second electrode pair sensing heart signal, wherein determine mate mark comprise along described first sensing vector sum described in second sensing vector in one or two sensing heart signal, and wherein said first sensing vector correspond to far field sensing vector and described second sensing vector correspond near field sensing vector.

3. equipment as claimed in claim 2, is characterized in that, described coupling mark determines in response to described far field sensing vector and described heart beating characteristic response is determined in described near field sensing vector.

4. equipment as claimed in claim 1, is characterized in that, described processor is configured to perform in response to determined coupling mark and determined heart beating feature increase or reduce one of event counter value.

5. equipment as claimed in claim 2, it is characterized in that, described processor is configured to further:

Determine described coupling mark whether in one of the first Matching band and the second Matching band corresponding to the second cardiac event that correspond to the first cardiac event;

Determine determined heart beating feature whether in one of described first Matching band and described second Matching band;

In response to described coupling mark in described first Matching band and determined heart beating feature in described first Matching band, by event counter value reduce first value; And

In response to described coupling mark in described first Matching band and determined heart beating feature in described second Matching band, event counter value is reduced and is less than the second value of the first value.

6. equipment as claimed in claim 2, it is characterized in that, described processor is configured to further:

Determine described coupling mark whether correspond to the first cardiac event the first Matching band, corresponding in one of second Matching band and the 3rd Matching band corresponding to the second cardiac event different from described first cardiac event of the first cardiac event; With

Determine determined heart beating feature whether in described first Matching band, one of described second Matching band and described 3rd Matching band, wherein said first Matching band have to have to the first dependency of untreatable cardiac event, described second Matching band than with the first dependency of untreatable event little with the second dependency of described untreatable cardiac event and the 3rd Matching band with can to treat cardiac event relevant.

7. equipment as claimed in claim 6, it is characterized in that, described processor is configured to further:

In response to described first coupling mark in described second Matching band and described heart beating feature in described 3rd Matching band, increase event counter value; And

In response to described first coupling mark in described second Matching band and described heart beating feature in one of described first Matching band and the second Matching band, reduce event counter value.

8. equipment as claimed in claim 7, is characterized in that, described untreatable cardiac event corresponds to supraventricular tachycardia event and described cardiac event for the treatment of corresponds to ventricular tachycardia.

9. equipment as claimed in claim 2, it is characterized in that, described processor is configured to further:

Determine whether described coupling mark is corresponding in one of the first Matching band of the first cardiac event, the second Matching band, the 3rd Matching band corresponding to the second cardiac event different from described first cardiac event and the 4th Matching band corresponding to the second cardiac event corresponding to the first cardiac event; And

Determine that whether determined heart beating feature is at described first Matching band, described second Matching band, described 3rd Matching band, with in one of described 4th Matching band, wherein said first Matching band has and the first dependency can treating cardiac event, described second Matching band has than little of described second dependency for the treatment of cardiac event with the first dependency can treating event, described 3rd Matching band has the first dependency with untreatable cardiac event, and described 4th Matching band have than with the first dependency of untreatable event large with the second dependency of untreatable cardiac event.

10. equipment as claimed in claim 9, it is characterized in that, described processor is configured to further:

In response to described first coupling mark in described second Matching band and described heart beating feature in described 4th Matching band, reduce event counter value; And

In response to described first coupling mark in described second Matching band and described heart beating feature not in described 4th Matching band, increase event counter value.

11. equipment as claimed in claim 9, it is characterized in that, described processor is configured to further:

In response to described first coupling mark at described first Matching band and described second coupling mark in described first Matching band, by event counter value increase by first value; And

In response to described first coupling mark at described first Matching band and described second coupling mark in described second Matching band, one of described 3rd Matching band and described 4th Matching band, described event counter value increase is less than the second value of described first value.

12. equipment as claimed in claim 2, it is characterized in that, described processor is configured to further:

Determine the first coupling mark corresponding to the described cardiac event sensed along described first sensing vector;

Determine whether described first coupling mark is corresponding in one of the first Matching band of the first cardiac event, the second Matching band, the 3rd Matching band corresponding to the second cardiac event different from described first cardiac event and the 4th Matching band corresponding to the second cardiac event corresponding to the first cardiac event;

Determine the second coupling mark corresponding to the described cardiac event sensed along described second sensing vector;

Determine that described second coupling mark is whether in described first Matching band, described second Matching band, one of described 3rd Matching band and described 4th Matching band;

Determine that whether determined heart beating feature is at described first Matching band, described second Matching band, described 3rd Matching band, with in one of described 4th Matching band, wherein said first Matching band has and the first dependency can treating cardiac event, described second Matching band has than little of described second dependency for the treatment of cardiac event with the first dependency can treating event, described 3rd Matching band is relevant to untreatable cardiac event, and described 4th Matching band have than with the first dependency of untreatable event large with the second dependency of described untreatable cardiac event,

In response to described first coupling mark in described second Matching band and correspond to described first sensing vector described heart beating feature in described 4th Matching band, by event counter value reduce first value;

In response to described first coupling mark in described second Matching band and the described heart beating feature corresponding to described second sensing vector in described 4th Matching band, event counter value is reduced and is greater than the second value of described first value; And