JP3338049B2 - Electrocardiogram signal analyzer and cardiac defibrillator including the same - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION CROSS REFERENCE TO RELATED APPLICATIONS This application is a continuation-in-part of Serial No. 541,881, filed on June 20, 1990, in the name of the same inventor and filed under the same name. is there.

TECHNICAL FIELD The present invention relates to an electrocardiogram signal analyzer, and more particularly to an electrocardiogram signal analyzer capable of detecting a heart disease, in particular, arrhythmia, fibrillation and related diseases, and a heart defibrillator including the same.

BACKGROUND OF THE INVENTION Despite significant advances in the diagnosis and treatment of ischemic heart disease over the last decade, a significant number of patients each year experience sudden cardiac death from ventricular fibrillation (VF). to date,
No reliable predictive or preventive measures have been developed. Externally, VF looks extremely complex and accidental. The same is true of other related heart diseases, including those in cardiac function, especially those that precede VF (signs of VF). Therefore, it is difficult to detect with certainty that a patient has VF or signs of VF with an automatic device. In addition, the signs of VF are:
It is difficult for a skilled medical person to make a reliable determination.

Thus, methods for detecting and assessing heart disease have wide applicability and utility. The patient monitoring device is used to
May call a healthcare professional. Automatic devices facing VF, such as implantable automatic defibrillators (AlCDs)
May change its behavior based on an assessment of the severity of the patient's condition. Also, a method of reliably assessing the risk of VF may have important utility in monitoring patients undergoing surgery or other critical care.

Certain antiarrhythmic drugs have been found to have proarrhythmic effects at excessive concentrations. For example, quinidine is known to have such toxicity. Methods for detecting and assessing heart disease also have broad applicability and utility in determining whether a patient is receiving a toxic (or semi-toxic) amount of a drug associated with a cardiac condition.

Chaos theory is a recently developed field related to phenomena that are highly complex and seemingly random, but can be described as deterministic consequences of relatively simple systems. Chaos theory has potentially wide application in ecosystems and other systems that include ambiguity and ambiguity. For example, chaos theory has been speculated to be valuable in describing certain natural processes, including electroencephalogram (EEG) and electrocardiogram (EKG) signals. Techniques for detecting and evaluating aspects of deterministic chaos are known in the field of chaos theory, but have few applications in the medical field.

Accordingly, there is a need for improved methods and devices for detecting and assessing heart disease, including ventricular fibrillation (VF) and signs of VF.

DISCLOSURE OF THE INVENTION A first feature of the present invention is a phase-plane plot of a patient's electrocardiogram (EKG) as a means for detecting heart disease.
t) An ECG signal analyzer for analyzing (PPP) is provided. Normal patients have a relatively smooth PPP. Patients who are at risk of developing signs of ventricular fibrillation (VF) have PPPs that exhibit characteristics of chaos, such as multiple zoning, "ban zones," periodic and phasic including cycle doubling, and VF The patient presenting has a PPP that looks noisy and irregular. The different PPPs are easily recognized and can thus detect patients with heart disease.

In a preferred embodiment, the degree of PPP of deterministic chaos can be measured by a processor, for example, by graph analysis and numerical analysis. (1) The processor may measure the Lyapunov exponent or fractal dimension of the PPP. (2) The processor may determine the Poincare section of the PPP and examine the Poincare section for signs of deterministic chaos. or,
The processed PPP and Poincare sections may be inspected by a human operator.

A second feature of the present invention is that a means for detecting heart disease is a frequency-domain transform of a patient's EKG.
The present invention provides an electrocardiogram signal analyzer for performing a ransform (eg, FFT) test. Normal patients have an FFT with a discontinuous spectrum, while patients exhibiting VF have a relatively continuous spectrum and relatively low frequencies (eg, about 5-6H).
z) has an FFT with peak energy. Patients with a VF that is difficult to recover from shock have an FFT with peak energy at relatively high frequencies (eg, about 10 Hz or higher).

In a preferred embodiment, the defibrillator may include means for providing a variable shock at least in part determined by the peak energy of the FFT. Also,
The defibrillator may include a means for alerting when the peak energy of the FFT is at a relatively high frequency.

A third aspect of the invention is the details of an action potential duration (APD) recovery curve or an action potential amplitude (APA) curve generated for a patient, for example, by fitting an exponential relationship to the curve or the parameter time constant of the curve. And a method for detecting drug toxicity based on the The slope of the fitted curve indicates the patient's potential for arrhythmia predisposition. Due to the difference in the parameters of the fit curve,
For example, normal and abnormal patients can be distinguished as to whether there is a risk of arrhythmia or ischemia. Normal patients have relatively low parameter time constants, and patients exhibiting drug toxicity have relatively high parameter time constants. APD or APA
The analytical techniques described herein may be used to generate a PPP of the data, interpret the PPP, and detect and evaluate drug toxicity.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 shows a patient monitoring device.

 FIG. 2 shows a series of sample EKG signals.

FIG. 3 shows a series of corresponding PPPs for the sample EKG signal of FIG.

 FIG. 4 shows a cross section of the sample PPP and the corresponding Poincare.

FIG. 5 shows a sample PPP and a corresponding Poincare section with time.

FIG. 6 shows a series of corresponding frequency domain transforms obtained by performing a fast Fourier transform (FFT) on the EKG signal.

FIG. 7 shows an improved implantable cardioverter defibrillator (AlCD).

FIG. 8 shows the signal response to stimulation of individual cardiac muscle cells, known in the art as action potentials.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The first aspect of the present invention relates to the detection and evaluation of heart disease by examining a phase diagram (PPP) of a patient electrocardiogram (EKG).

FIG. 1 shows a patient monitoring device. Patient 101 has an electrocardiogram (E
KG) coupled to a device 102, which acquires EKG signals and sends them to a processor 103. The processor 103 may display the EKG signal on the monitor 104 as is well known in the art, or may process the EKG signal and display the processing result on the monitor 104.

EKG signals are well known in the art, as are the methods of obtaining them. EK as used here
G refers to a surface electrocardiogram, but other types of electrocardiograms can be used with the methods disclosed herein and are within the scope and spirit of the invention. For example, the EKG mentioned here has a surface E
KG, epicardial EKG, endocardial EKG, or other relevant signal (or set of signals) measured in or near the heart.
Further, the signals handled may be voltage signals, current signals or other related electromagnetic values (or series of values).

FIG. 2 shows a series of sample EKG signals. The first EKG signal 201 indicates a normal patient. The second EKG signal 202 indicates a patient in transition to VF. Third EKG signal 203 indicates a patient suffering from VF.

Processor 103 may construct a phase plane diagram (PPP) from the EKG signal. The first type of PPP has E for its first derivative
This is a plot of the KG variable. In a preferred embodiment,
The EKG variable is the voltage v (a function of time) and its first derivative is dv / dt (also a function of time).

However, upon perusal of the specification, drawings and claims, it will be apparent to those skilled in the art that the permissiveness of PPP interpretation is wide. The variables chosen for the PPP can be a variety of different parameters, including EKG voltage, current, or other signal values. The chosen variable (v) is its first derivative (dv / dt), its second derivative d ² v / dt ² ,
Or, it may be plotted against another nth derivative d ⁿ v / dt ⁿ . Alternatively, the Mth derivative may be plotted against the Nth derivative.

Other forms of PPP use their own time-delay conversion (time de
(e.g., v (t) versus v (t- [delta] t)) for a plot of the EKG variable (or its Nth derivative) against the laid version. This form of PPP is sometimes called a "return diagram."
This type of PPP is not sensitive to EKG signal noise.

Another type of PPP is that three EKG variables (or their Nth
(For example, v, dv / dt and d ² v / dt ² ). Such a PPP is three-dimensional. If the PPP is three-dimensional, it may be displayed in three dimensions,
The two-dimensional surface “cut” of the three-dimensional display may be displayed on the two-dimensional display. It will be apparent to one skilled in the art that all of these alternatives or combinations thereof described herein are feasible and within the scope and spirit of the present invention.

FIG. 3 shows the corresponding PPP group for the sample EKG signal of FIG. The first PPP 301 corresponds to the first EKG signal 201. Second PPP3
02 corresponds to the second EKG signal 202. The third PPP303 is the third EKG
Corresponds to signal 203.

Part of this feature of the invention is that the normal patient has a PPP that indicates the regularity and smoothness of the EKG signal from the normal patient,
On the other hand, patients receiving VF have deterministic chaos (eg,
EKG for period, banding and "ban zone")
It is the discovery of having a PPP that indicates signal irregularity and complexity. In addition, patients undergoing a transition from normal to VF (ie, during the onset of VF) exhibit PPP consistent with their assessment that their EKG signal is undergoing a transition to deterministic chaos.

Normal patients have a relatively regular beat-to-beat EKG
Signal. When the patient switches to VF, the patient's EKG signal initially exhibits paired degeneration and variations between regular beat-to-beat signals. As the conversion continues, the patient's EKG signal will show an ever increasing number of variables and fluctuations between the regular signals (eg, 4 alternating beats, 8 alternating beats, etc.), and by the end, no more degenerate and regular signals The EKG signal becomes irregular and extremely complex, which cannot be confirmed. At this point, the patient has
Is generally referred to as

Similarly, the patient's PPP changes from a smooth single band display to a multiple band display (indicating multiple alternating beats) and finally to an irregular high complex display. The display changes in PPP are so dramatic that even relatively untrained people can see the difference. This is significantly different from the display changes in the EKG that generally require a skilled cardiologist to evaluate.

There are several possible factors that change a patient from normal to VF. These factors can include drug overdose (especially overdose with antiarrhythmic drugs that have proarrhythmic effects at overdose, such as quinidine intoxication), hyperelectric stimulation, cooling, ischemia and stress. In a preferred embodiment, the patient monitor may examine the patient's PPP to determine if the patient has changed from normal to VF. This indicates that these proarrhythmic factors are excessively present.

Processor 103 may further process the PPP to determine the degree of deterministic chaotic PPP. Several techniques can be applied for this purpose.

(1) The processor 103 can measure the Lyapunov exponent of PPP. The Lyapunov exponent for PPP is a measure of the degree to which roads near PPP diverge. The Lyapunov exponent is well known in chaos theory and can be measured with available software. See, for example, Wolff et al., "Determining Lyapunov Exponents from a Time Series", Fujika Day 1985; 16: 285-317.

(2) The processor 103 can measure the fractal dimension of the PPP. The fractal dimension of PPP is a measure of the degree to which PPP forms a "space-filling" curve. Fractal dimensions are well known in chaos theory and can be measured by several techniques (eg, correlation dimensions or box calculations), for example, as shown below.

Processor 1 to measure the fractal dimension of PPP
03 prints a straight grid (consisting of a series of boxes) on the PPP and counts the number of boxes cut by the PPP trace. Processor 103 changes the size of the grid,
Record the size and total number of each grid. Next, the processor 103 calculates a constant k according to the following relationship.

ln (# (number) of box cutting) = k * ln (# (number) of box in grid 304) ln (# of box cutting cut) = k * ln (# of box in grid 304) The constant k is It is a measure of the fractal dimension of PPP. About 3
The value of k with a fractional part between and about 7 is PPP
Would represent a deterministic chaos-based process, and thus a patient near VF (ie, a patient who is substantially VF).

(3) Processor 103 may measure the Poincare section of the PPP and examine the Poincare section for an indicator of deterministic chaos as described herein. The processed PPP and Poincare sections are also shown for inspection by a human operator, so that all visible structures are easily recognized.

FIG. 4 shows a sample PPP 401 and a corresponding Poincare section 402. The Poincare section includes a line drawn across a portion of the PPP. Generally, such a line segment is
It is almost perpendicular to the trajectory of PPP.

Processor 103 acquires data points at each Poincare section or PPP and calculates a statistical measure of the isotropic or non-uniformity of these data points. Such indicators are based on the mean and standard deviation of these data points (these can be calculated by statistical methods well known in the art). Ratio γ
Γ = (standard deviation) / (expected value) (403) is a measure of the degree of aggregation in the Poincare section.

Larger values for γ tend to represent a process where PPP is based on deterministic chaos, and thus patients who are close to VF (or who are actually VF). The value for γ, as is well known in the art, is compared with the value for γ for a normal patient, along with a set of credit bands to indicate the degree of change from a normal patient, for testing by a human operator. Is shown in

Processor 103 may also calculate other statistical indicators of the Poincare section.

Processor 103 may also inspect the "aged" Poincare section of the PPP.

FIG. 5 shows the embodiment PPP501 and the corresponding aging Poincare section 502
Is shown. The Poincare section over time may include a PPP selected set of data points by selecting one data point every t seconds.
Aged Poincare sections can be analyzed in a manner similar to the other Poincare sections disclosed herein.

A second aspect of the invention relates to the detection and evaluation of heart disease based on a frequency domain transform of a patient EKG.

FIG. 6 shows a corresponding frequency domain transform group obtained by performing an FFT on the EKG signal. The first transform 601 corresponds to a first EKG signal (not shown). Second transform 602 corresponds to a second EKG signal (not shown).

In a first transform 601 representing a normal patient, the frequency spectrum indicates that the energy of the corresponding EKG signal occurs primarily at the frequency of the discontinuous location. In a second transform 602 representing a patient exhibiting VF, the frequency spectrum indicates that the energy of the corresponding EKG signal has a continuous spectrum of frequencies and has an energy peak 603.

Part of this feature of the invention is that the harmonic amplitude ratio (HM
R) the use of visual and mathematical methods to analyze the frequency domain transform, including the calculation of To measure the HMR, the main peak or central region of the energy distribution in the spectrum of the frequency domain transform (eg, FFT) was specified, and the HMR was calculated as follows. The magnitude of the transformation in the region of the particular point is determined, for example, by summing the magnitude of the transformation at the particular point and the surrounding points, and the corresponding magnitude in the region of the harmonic value of the frequency for the particular point. And is summed up. This sum is divided by the total magnitude of the transform for the entire signal. The ratio is defined as HMR.

One known method for extracting VF from a patient ("defibrillation") is to apply electrical stimulation to the patient's heart. This electrical stimulus generally has substantial energy, for example, 10
It has -20 joules and often causes tissue damage to patients, even if the patient is successfully defibrillated. Generally, a number of stimuli that increase energy are required. Therefore, it is advantageous to use large stimuli only when needed,
It is advantageous to use as few stimuli as possible. Part of this feature of the invention is the discovery that relatively low energy stimulation is generally sufficient to defibrillate a patient when the energy peak 603 of the frequency domain transform 602 is at a relatively low frequency. When the energy peak 603 of the frequency domain transform 602 is at a relatively high frequency (and when the second energy peak 604 appears at the frequency domain transform 602 at a relatively high frequency), the patient may be defibrillated by electrical stimulation anyway. Defibrillating the patient, if possible, requires relatively high energy stimulation.

One application of this finding is in an automatic implantable cardioverter defibrillator (AlCD) that attempts to automatically detect VF and automatically apply a stimulus to defibrillate the patient.

FIG. 7 shows the modified AlCD701. The patient 702 acquires the EKG signals and communicates them to an AlCD processor 704 that adjusts a stimulator 705 to deliver defibrillation stimulation to the patient 702.
Connected to EKG703.

The improved AlCD701 is also available (eg, AlCD processor 7
Includes software (as part 04) for measuring the FFT of the EKG signal and for measuring the energy peaks in the FFT. When the energy peak in the FFT is relatively low, AlCD processor 704 adjusts stimulator 705 to provide a relatively small stimulus to the patient. When the energy peak in the FFT is relatively high, the AlCD processor 704 adjusts the stimulator 705 to provide a relatively large stimulus to the patient,
Also, an alert 706 or other indicator that defibrillation was not successful may be sent.

A third aspect of the invention is the detection and evaluation of drug toxicity based on a parameter time constant for an action potential duration (APD) recovery curve or an action potential amplitude (APA) curve configured for a patient. About.

FIG. 8 shows the signal response to stimulation of individual cardiac muscle cells. This individual cellular response is known in the art as "action potential".

It is well known in the art that the time duration for the recovery 801 of individual cells depends on factors including the rest period 802 that the cells have before stimulation. It is also well known in the art that APD return curves can be configured for human patients with the use of intracardiac catheters. However, the full relationship between the actual time duration for the withdrawal 801 based on the rest period 802 is not known.

Part of this feature of the invention is based on the finding that when the time duration for recovery 801 is plotted against rest period 802 (diastolic interval), the curve follows an exponential relationship.

APD = APD _pl −A * exp (−DI / tau) (803) where APD _pl is plateau-APD. A is a balancing constant. DI is the diastolic interval and tau is the parameter time constant.

The non-linear nature of the APD recovery curve can promote deterministic chaos in response to overstimulation of cardiac muscle cells. The steeper the APD recovery curve (ie, the greater the parameter time constant), the greater the preference of the heart to enter the VF. That is, another part of this aspect of the invention is that normal patients have relatively low APD recovery parameter time constants, while patients exhibiting drug toxicity (eg, quinidine intoxication) have relatively high APD recovery parameters. It is the discovery that it has a time constant. Recovery parameter time constants can also be used to monitor cardiac stability and to assess the efficacy of antiarrhythmic drugs.

 The experimental transformation of the present invention is achieved by the present inventors.

Experiment 1 The mathematical study used PPPs, return maps, Poincare sections, correlation dimensions and spectral analysis to distinguish periodic, chaotic, and arbitrary signals. PPPs were useful in distinguishing between all three classes of signals. The periodic signal showed a clear and widely separated trajectory. The chaotic signal showed banding, forbidden areas and a sharp initial condition dependence. Any signal did not show any obvious internal structure. Excluding the noise effect, the major difference between the PPPs and the appropriately delayed recovery map was a 45 degree rotation. Poincare sections could also distinguish between the three classes of signals. The periodic signal showed a separation point. The chaotic signal showed a clear self-similar ordered area. The arbitrary signal showed a Gaussian distribution of points. The correlation dimension distinguished more between chaos and arbitrary signals than between chaos and periodic signals. Spectral analysis using FFTs and harmonic intensity ratios (HMR) was able to distinguish between periodic signals, but not between arbitrary and chaotic signals. HMRs of periodic signals were greater than 97%.

HMRs of chaotic signals varied between 17 and 80%. The HMRs of the arbitrary signal were about 40%. PPP was heavily affected by noise. The recovery map was hardly affected. On the other hand, spectral analysis was relatively immune from noise. PPPs,
It was concluded that the recovery map, Poincare section, correlation dimension and spectral analysis were all useful determinants of the chaotic system.

Experiment 2 Mathematical studies have particularly focused on the possibility of spectral analysis to distinguish chaos from arbitrary signals. In this experimental example, two series of arbitrary signals were generated. The first sequence contains 5000 pseudo-random numbers simplified using the method of least squares approximation. The second series contained white noise obtained from an analog-to-digital converter. Spectral analysis was performed by applying the FFT to the data and examining the change from a broad band spectrum or narrow band to a wide band which was assumed to be a judgment of chaos. It was concluded that spectral analysis by itself was insufficient to clearly distinguish chaotic signals from arbitrary signals, and that additional tests, such as PPP and recovery maps, were necessary for this purpose.

Experiment 3 The experimental example examined spectral analysis, visualization of PPPs, and correlation dimension analysis to help distinguish between normal curvature rhythm and VF in dogs. Ischemia and reperfusion were used as stress factors in dogs with closed chest anesthesia. Spectral analysis of dogs with normal curvature rhythm revealed a narrow band spectrum with a fundamental ratio at the curvature ratio and harmonics extending beyond 50 Hz. PPPs were consistent with the periodic dynamics, and dimensional analysis revealed a lower dimensional response (1--2.5). Symmetrically, VF
Spectral analysis of dogs with a. Revealed a broad band response with most of the energy at 6 Hz and with energy at all frequencies between 1 and 25 Hz. PPP shows a forced aperiodic response, and dimensional analysis shows higher dimensions (4-6) than observed for dogs with normal curved rhythms
Revealed. That is, all three techniques proved to be useful for distinguishing normal curvature rhythm from VF.

Experiment 4 The experiments examined the spectral analysis, the visualization of PPPs, the visualization of the recovery map, and the correlation dimension, which are useful for confirming VF in humans. These analysis techniques were applied to data from eight hypothermic patients undergoing spontaneous VF and from three normal body temperature patients with VF induced during electrophysiology. All patients have a broad band frequency spectrum (0-12 Hz), low dimension (range 2-
5) and had banned areas with PPPs and recovery map belts. It was concluded that spectral analysis, visualization of PPPs, visualization of recovery maps and correlation dimension analysis were useful for detecting and evaluating VF.

Experiment 5 The experiment examined spectral analysis, visualization of PPPs and correlation dimension analysis, which was useful in distinguishing between normal curvature rhythm and VF in humans. 8 undergoing open heart surgery
VF in human hypothermic human patients was studied. In all patients, first and second order PPPs indicate forbidden areas and zoning, and FFTs range from 0 to 25 Hz, often with powers below 12 Hz.
Showed a relatively continuous power spectrum at all frequencies. Symmetrically, in all cases, the correlation dimension was less than 4. It was concluded that multifaceted analysis of the data (= spectral analysis + correlation dimension analysis) is more desirable than relying on a single analysis technique such as correlation dimension.

Experiment 6 The experiment used spectral analysis and visualization of PPPs to explain the heterogeneous nature of atrial fibrillation. In the experiment,
The examiner caused acute fibrillation by a rapid procedure of stimulation of the atria of seven closed-chest dogs. PPPs based on EKG data often fill well-defined structures, and digitized EKG FFTs typically have a continuous spectrum that is distinct from apparent harmonic components or altered in a time and place dependent manner, typically at 15 Hz. The following peaks were shown. It was concluded that spectral analysis and visualization of PPP were useful techniques for analyzing atrial and ventricular fibrillation.

Experiment 7 In experiments, visual analysis of PPPs and the slope of the APD return curve proved to be useful for detecting and assessing quinidine-induced VF in in vivo hearts. Quinidine is 90-10
All doses were administered at 0 mg / kg or 5 minutes at 30 minute intervals until ventricular tachycardia or VF occurred, whichever came first. The PPPs of quinidine intoxicated cells showed a sharp initial condition dependence and the presence of a forbidden zone, and the corresponding FFTs showed a continuous spectrum. In contrast, the PPPs of cells in control dogs were homogeneous and densely packed, and the corresponding FFTs displayed distinct spectra. The initial slope of the APD recovery curve for quinidine-excited cells was at least as large as that of normal cells. It was concluded that quinidine toxicity was associated with the slope of the APD recovery curve.

Experiment 8 The experiment compared the slope of the APD and APA recovery curves with quinidine intoxication. Quinidine was administered to eight dogs for 5 hours (90-100 mg / kg). Three untreated dogs served as controls. Ventricular and Purkinje cells from treated and untreated dogs were then subjected to electrical stimulation on a cycle of 900 to 600 msec. Reducing the cycle length to 600 msec resulted in APD and APA irregular dynamics, including electrical alternating beats and bifurcations. Calculate the slope of the APD recovery curve,
Quinidine for Purkinje fibers and ventricular muscle cells
It was found to be steeper in excitable cells than during quinidine efflux or in normal untreated cells. The curves were matched by the exponential equation given here. APA changes almost always correlated with APD changes. In the three normal tissue preparations, neither ventricular muscle cells nor Purkinje cells showed a divergent response for APD or AA. Quinidine toxicity, and possibly other pre-drug-induced arrhythmias, is
It was concluded that both D and APA correlated with the slope of the recovery curves.

Experiment 9 In the experiment, quinidine-induced ventricular palpitations and VF were measured in dogs with action potential duration (APD).
ration) and PPP generated from action potential amplitude (APA) data
Analysis was performed using s. Both PPPs showed a ban zone and a sharp initial condition dependence, suggestive of chaos. APD or APA
It was concluded that the PPPs based on PEG were useful for detecting and assessing quinidine toxicity.

Experiment 10 In the experiment, EKGs of quinidine-excited dogs were analyzed by frequency spectrum, phase diagram, Poincare section, recovery map and Lyapunov exponent. In control conditions and at therapeutic doses, PPPs are uniformly thick and cycle-to-
It showed no gap indicating that the cycle change was due to normal biological "noise".

However, when the quinidine dose was increased to intermediate levels (40-50 mg / kg), PPPs showed uneven thicknesses indicating a sharp initial condition dependence and were also marked by distinct bands (boundaries or gaps). (Dark black coloring). At these intermediate doses, the Lyapunov index became positive and the Poincare recovery map also showed non-arbitrary chaos. PP at higher dose
Ps has become more complex. In the two dogs that showed VF (and not elsewhere), there was a clear change in the last pre-fibrillation dose in PPP, development of a "funnel", a standard mechanism of chaos. At all pre-fibrillation doses, the frequency spectrum was distinct with peaks at the fundamental frequency and polyharmonic. Chaotic doses occurred during progressive quinidine intoxication, and it was concluded that PPPs and graphs and numerical analyzes based on PPPs were better indicators of chaos than frequency spectra.

Experiment 11 In the experiment, quinidine poisoning in dogs was assessed by APA and APD.
Analysis was performed using PPPs generated from the data. EKG record is 100
The experiment was performed at various driving rates from 0 to 560 msec. The increase in transmission rate from 1000 to 500 msec caused a progressive appearance of higher order periodicities (periods 3 and 4). Phase fixation was seen in a stimulus (S) response (R) pattern that was periodically repeated for all four preparations at S: R ratios of 2: 1, 5: 3, 3: 2. Aperiodic changes in APA and APD were seen at higher transmission rates. A number of intermediate stages in predicting chaos were found in quinidine excitable fibers. These results further demonstrate the utility of the method of the present invention for detecting quinidine intoxication and precursor stages of intoxication.

Experiment 12 In the experiment, quinidine toxicity in dogs was determined by APA and APD.
Analysis was performed using PPPs generated from the data. 20 electrical stimuli
It was used to move heart tissue at various rates from 00 to 300 msec. These stimuli were 108 ± 36 msec and 12 ±
A stable alternating beat (branch) occurred at 9 millivolt APD and APA. More increase in transmission rate caused irregular dynamics. This transition preceded 50 consecutive beats with various repetitive stimulus-response ratios (phase-fixed). Such a motive was not attracted in the three untreated (control) tissues. The APD recovery curve had a significantly (P <0.05) steeper slope than the six control fibers.
The stimulus-response latency was still constant at 6-9 msec. PPPs of APDs during random dynamics showed a sharp initial condition dependence and forbidden zone consistent with chaos theory. These results demonstrate the utility of the method of the invention to detect precursors to quinidine addiction and toxicity.

Experiment 13 The experiment used spectral analysis, PPPs, Poincare section, Lyapunov exponent, and dimensional analysis to analyze computer simulated experimental waveforms including sine waves, modulated sine waves, square waves, sawtooth waves, and triangular waves. The inspector added arbitrary noise to the waveform at 1%, 10% and 20%. The experiments further show that VF involves five different involvements: quinidine intoxication, premature electrical stimulation following quinidine intoxication, coronary obstruction, acute ischemic myocardium reperfusion, and EKG data from anesthetized dogs caused by global hypothermia. The same analysis technique was used. The preliminary result is
PPPs and Poincare sections in dogs undergoing ventricular fibrillation were consistent with chaos, while spectral analysis showed no indication of chaos. The investigators concluded in part that VF could be described as a chaotic electrophysiological function, but no single method of analysis was sufficient to detect such function.

One conclusion that can be drawn from the tests cited herein is that the analysis of each of the features of the present invention is enhanced in combination with one or more of the other features of the present invention. Preferred embodiments of the invention may include a combination of the features of the invention described herein. One preferred embodiment is
Multi-dimensional analysis of PPP (eg, visually on display,
Graphically in Poincare section and numerically in Lyapunov exponent and correlation dimension), frequency spectrum analysis and AP
It may include a mathematical analysis of the D recovery curve.

Other Embodiments While preferred embodiments are described herein, many transformations are possible, which are within the spirit and scope of the invention, and which transformations may occur after perusal of the specification, drawings and claims. Will be apparent to those skilled in the art.

It should also be appreciated that embodiments of the present invention may include means for continuous monitoring of drug toxicity, atrial fibrillation, ischemia or other cardiac conditions, such as surgery or patient recovery from surgery. It will be clear to the skilled person. Further embodiments of the present invention may include means for indicating the detected cardiac condition to the attending physician medical staff or patient.
In a preferred embodiment of the invention, a means is provided for the patient to contact a physician (when a heart condition is detected) or to go to a nearby hospital for treatment.

──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Diamond, George Alexander United States 90068 California, Los Angeles, Wild Oak Drive No. 2408 (72) Inventor Can, Steven Shade United States 90064 California, Los Angeles, Manning Avenue 2241 (72) Inventor Denton, Timothy Alan 90048 California, Los Angeles, Norwich Drive 513 (72) Inventor Evans, Steven United States 20022 New York, New York, East Fifty Second Street No. 325 (56) References JP-A-63-203133 (JP, A) JP-A-5 9-189831 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) A61B 5/0452-5/0472 A61B 10/00

Claims (36)

(57) [Claims] 1. An electrocardiogram signal receiving means, a signal processing means for processing the electrocardiogram signal to determine a phase diagram of the electrocardiogram signal, and a display means for displaying the phase diagram.
The signal processing means superimposes a straight grid constituting a large number of boxes on the phase diagram, calculates the number of boxes cut by traces of the phase diagram, and calculates the number of boxes in the grid. The fractal dimension of the phase diagram is determined by calculating a constant k, which is a measure of the fractal dimension, by the formula ln (number of cut boxes #) = k * ln (number of boxes in the grid). k is 3
An electrocardiogram signal analysis device, which determines whether or not it is within a range of ~ 7. 2. An electrocardiogram signal analyzer according to claim 1, wherein said phase diagram comprises a multidimensional diagram of at least two variables. 3. An electrocardiogram signal analyzer according to claim 1, wherein said phase diagram comprises a multidimensional diagram of at least three variables. 4. The electrocardiogram signal analyzer according to claim 1, wherein said phase diagram plots said signal voltage against its derivative. 5. The derivative of claim 4, wherein said derivative is a first derivative.
An electrocardiogram signal analyzer according to claim 1. 6. The fractal dimension changes the size of the grid, each time recording the size of each box and the number of boxes cut by the trace of the topological diagram as well as the number of boxes in the grid. The electrocardiogram signal analyzer according to claim 1, wherein the electrocardiogram signal analyzer is determined by calculating a constant k, which is a measure of the fractal dimension of the phase diagram. 7. An electrocardiogram signal receiving means, a signal processing means for processing the electrocardiogram signal to determine a phase diagram of the electrocardiogram signal, and a display means for displaying the phase diagram.
An electrocardiogram signal analyzer, wherein a Lyapunov exponent of the phase diagram is determined. 8. An electrocardiogram signal receiving means, signal processing means for processing the electrocardiogram signal to determine a phase diagram of the electrocardiogram signal, and display means for displaying the phase diagram.
The signal processing means calculates an intersection of a phase diagram and a two-dimensional section, and determines a Poincare section for an index of deterministic chaos by acquiring a series of data points at the intersection. Electrocardiogram signal analyzer. 9. The electrocardiogram signal analyzer according to claim 8, wherein the two-dimensional cross section is a line segment. 10. The electrocardiogram signal analyzer according to claim 8, wherein said signal processing means calculates anisotropy or heterogeneity of said series of data points. 11. An electrocardiogram signal analyzer according to claim 8, wherein said signal processing means calculates a ratio of an average deviation and a standard deviation of said series of data points. 12. The electrocardiogram signal analyzer according to claim 8, wherein said signal processing means calculates a statistical index of said series of data points and causes said display to display said statistical index for an operator's examination. 13. A means for receiving an electrocardiogram signal, and a signal processing means for determining a phase plane diagram of the electrocardiogram signal for examining a Poincare section for an indicator of deterministic chaos, and determining a Poincare section of the phase diagram. And an electrocardiogram signal analyzer comprising: a display unit for displaying the phase diagram and the Poincare section. 14. The electrocardiogram signal analyzer according to claim 13, wherein the phase diagram comprises a multidimensional diagram of at least two variables. 15. The electrocardiogram signal analyzer according to claim 13, wherein the phase diagram comprises a multidimensional diagram of at least three variables. 16. The electrocardiogram signal analyzer according to claim 13, wherein the phase diagram plots the signal voltage against its derivative. 17. The derivative of claim 1, wherein said derivative is a first derivative.
16. The electrocardiogram signal analyzer according to 16. 18. A signal processing means for acquiring a data point in a Poincare section or a phase diagram and calculating a statistical index of anisotropy or inhomogeneity of the data point.
14. The electrocardiogram signal analyzer according to 13. 19. The electrocardiogram signal analyzer according to claim 18, wherein the index is based on an average and a standard deviation of the data points. 20. A ratio (r) between a standard deviation, which is an index of the degree of cohesion of the Poincare section, and an expected value,
19. The electrocardiogram signal analyzer according to claim 18, wherein the r value is displayed by a display means together with the confidence band group. 21. Means for receiving an electrocardiogram signal from an electrocardiogram device, defibrillation shock applying means for outputting a defibrillation shock, and processing means for processing the electrocardiogram signal and controlling the defibrillation shock applying means. The processing means may provide the defibrillation shock applying means with a shock having a relatively large energy when the frequency domain transformation has an energy peak at a relatively high frequency, and the frequency domain transformation has an energy peak at a relatively high frequency The automatic implantable cardioverter defibrillator, characterized in that it outputs the following. 22. The apparatus according to claim 21, wherein said signal processing means performs a Fourier transform. 23. The signal processing means for determining whether the frequency domain transform includes a continuous spectrum.
An apparatus according to claim 1. 24. The apparatus of claim 21, wherein said signal processing means determines whether said frequency domain transform includes an energy peak at a relatively high frequency. 25. The apparatus according to claim 24, wherein said relatively high frequency is higher than about 6 Hz. 26. The apparatus of claim 24, wherein said relatively high frequency is higher than about 10 Hz. 27. A means for receiving an electrocardiogram signal from an electrocardiogram device, a signal processing means for establishing a relationship between a diastolic interval and an action potential duration from said electrocardiogram signal and determining a time constant value of said relationship. An electrocardiogram signal analyzer for detecting drug toxicity, comprising: means for displaying a constant value. 28. The electrocardiogram signal analyzer according to claim 27, wherein said signal processing means determines whether said time constant value is substantially equal to or greater than a normal value. 29. An electrocardiogram signal analyzer according to claim 27, wherein said signal processing means calculates a Poincare section, a Lyapunov exponent, or a correlation dimension of a phase diagram. 30. A means for receiving an electrocardiogram signal, plotting a relationship between a signal voltage of the electrocardiogram signal and a derivative of the signal voltage, determining a phase diagram of the electrocardiogram signal according to the relationship, and determining the phase diagram. An electrocardiogram signal analyzer comprising: signal processing means for determining a Poincare section in the figure; and display means for displaying the Poincare section. 31. The signal processing device according to claim 30, wherein the signal processing means calculates an intersection between the phase diagram and the two-dimensional section, and determines a Poincare section by acquiring data points at a series of intersections of the intersection. ECG signal analyzer. 32. The electrocardiogram signal analyzer according to claim 31, wherein the two-dimensional section is a line segment. 33. An electrocardiogram signal analyzer according to claim 30, wherein said signal processing means calculates anisotropy or inhomogeneity of said series of data points. 34. The electrocardiogram signal analyzer according to claim 30, wherein said signal processing means calculates a ratio between an average deviation and a standard deviation of said series of data points. 35. The electrocardiogram signal analyzer according to claim 30, wherein said signal processing means calculates a statistical index of said series of data points and displays said statistical index on a display device for an operator's examination. 36. The electrocardiogram signal analyzer according to claim 30, wherein said signal processing means determines a temporal Poincare section of said phase diagram.