WO2018073722A1

WO2018073722A1 - A computer implemented method to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue and computer programs thereof

Info

Publication number: WO2018073722A1
Application number: PCT/IB2017/056404
Authority: WO
Inventors: Antonio Berruezo Sánchez; Alejandro ALCAINE OTIN; Juan Pablo MARTÍNEZ CORTÉS; Pablo LAGUNA LASAOSA; Oscar CÁMARA REY; David SOTO IGLESIAS
Original assignee: Universidad de Zaragoza; Hospital Clinic de Barcelona; Universitat Pompeu Fabra UPF
Current assignee: Universidad de Zaragoza; Hospital Clinic de Barcelona; Universitat Pompeu Fabra UPF
Priority date: 2016-10-17
Filing date: 2017-10-16
Publication date: 2018-04-26
Anticipated expiration: 2019-04-17

Abstract

A computer implemented method to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue and computer program products A plurality of mapping points acquired from a patient are stored in a signal acquisition unit, said mapping points including a ECG signals, EGM signals and a 3D location of the EGM signals, the method comprising for a reference mapping point: a) detecting each beat present in one recorded ECG signal and identifying a beat of interest from the detected beats; b) identifying a principal EGM wave related to the identified beat of interest; c) identifying an onset and an end time landmarks of said principal EGM wave providing a primary delineated EGM signal and measuring a voltage amplitude of the primary delineated EGM signal; d) performing a further analysis of the primary delineated EGM; and e) creating a conducting channel map and a propagation map of the heart based on the result of a tagging performed during said analysis.

Description

A computer implemented method to identify the ventricular arrhythmoqenic substrate in myocardial scar or fibrotic tissue and computer programs thereof

Technical Field

The present invention relates in general to the field of medical signal processing techniques. In particular, the invention relates to a computer implemented method, and computer programs, to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue. To identify said ventricular arrhythmogenic substrate a qualitative analysis of the intracardiac electrogram (EGM) signal during normal sinus rhythm or during arrhythmia rhythm of a patient is performed to be later used as guidance for catheter ablation interventions of scar-related ventricular tachycardias (VTs).

Background of the Invention

The analysis of substrate voltage maps using an electroanatomical mapping (EAM) system is one of the most used techniques for mapping the arrhythmogenic substrate in scar-related VTs in order to identify the isthmuses of the VT and thus perform the ablation. The peak-to-peak voltage amplitude of the recorded EGMs can help to locate the small bundles of viable tissue within the scar (known as slow conducting channels (or CCs)) which form the substrate for reentry and promotion of VTs; therefore, those slow conducting channels become the area of interest of the ablation treatment. Current substrate mapping inaccurately identifies those areas of interest, forcing clinicians and companion technicians to manually perform this task, thus being affected by many variables dependent on the substrate itself, catheter used, operator expertise, etc., which increases its imprecision. This has led to think that a qualitative analysis (rather than quantitative) of the electrical signal, could identify more precisely the slow conducting channels and in addition, the most vulnerable parts of these or those parts allowing a more efficient ablation treatment. Electrically, EGM signals from slow conducting channels in the scar are characterized by the presence of very low amplitude and high-frequency delayed potentials reflecting the delayed conduction of the viable tissue within the myocardial scar. These delayed potentials are often preceded by a broader and lower frequency activation or "far-field" caused by the healthy tissue surrounding the scar area. It is common to find fibrotic or scar areas in the ventricles that are surrounded by healthy tissue (which is electrically activated producing signals of higher amplitudes), therefore, when the bipolar signal is acquired within the scar, this bipolar signal presents a component of large amplitude and in synchrony with the heartbeat-related contraction or surrounding healthy tissue activation (the "far-field" potential). These far-field potential may mask the existence of delayed potentials (near field or "near-field") of lower bipolar voltage (amplitude), which characterize the slow conduction channels. Above all, this masking is evident at the edges delineating the scar where healthy tissue and tissue "border zone" intermingle. That is why a simple measure like the present bipolar voltage is heavily influenced by the existence of far-field potentials in the measurement window.

As said, nowadays, the identification of the presence of delayed potentials is done manually by an operator of the electroanatomical mapping system assisted by the electrophysiologist in charge of handling the catheters and intervention. This identification is very subjective and dependent on the operator training and experience. Furthermore, this manual identification is more expensive when multipole mapping catheters are used due to the amount of signals acquired during a single cardiac cycle.

US-A1 -2016128785 discloses a system and a method for identifying the arrhythmogenic circuit of a patient or subject. In one embodiment, the method comprises obtaining data for electrograms recorded at various locations of the heart while programmed ventricular pacing with extra stimuli was performed, obtaining decrement values for at least two different locations of the heart using the recorded electrograms, generating at least a portion of a decrement map using the decrement values, and identifying the arrhythmogenic circuit based on electrograms having significant decremental properties. Unlike the proposed invention, the system, and corresponding method, of this patent application is not based on an electroanatomical mapping system but on a multipoint mapping system. Besides, the system cannot perform an automatic identification of the type of point (channel entrance or double potential) nor it performs a previous step of 'signal conditioning' using information from several reference-like heart beats to incorporate information from these in the same signal and thus be less dependent on one measure. US-A1 -2014235996 discloses a method for mapping of myocardial electric activity includes measuring electrocardiogram data or magnetocardiogram data and mapping the degree of electric activity of a myocardial surface using the electrocardiogram data or the magnetocardiogram data. A signal source of the electrocardiogram data or the magnetocardiogram data is a myocardial surface potential that is scalar quantity. The mapping uses a lead-field vector which represents the sensitivity between the myocardial surface potential and the electrocardiogram or magnetocardiogram data, and a modified lead-field vector which combines a constraint matrix with a constraint condition where no potential sources exist in a specific region. Unlike the proposed invention, in this patent application the slow conductive channels are not identified nor is a qualitative analysis associated with the existence or not or a delayed potential and its relation with a previous potential is performed. More adequate automatic processing methods are therefore needed to assist the catheter ablation treatment in patients with scar-related VTs due to slow conducting channels.

Description of the Invention

The proposed invention aims to solve the above-mentioned problems, by identifying the presence or absence of delayed potentials in the bipolar electrogram signal. To that end, a tool is provided automating and speeding up decision making during surgery, thus supporting the characterization of the slow conduction channels and identifying entries of such channels as targets for ablation. Furthermore, this identification can eliminate redundancies in the acquisition of similar and very close signals and can associate a correct measurement of the bipolar voltage (amplitude) to more accurately characterize the scar.

In addition, the new bipolar voltage maps are compared with contrast nuclear-magnetic resonances (delayed gadolinium) checking a better correlation of these voltage maps than that of the ones originally obtained.

Embodiments of the present invention provide according to a first aspect a computer implemented method to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue, wherein a plurality of mapping points acquired from a patient are stored in a signal acquisition unit, said plurality of mapping points including a plurality of electrocardiogram (ECG) signals, a plurality of corresponding intracardiac electrogram (EGM) signals, and a 3D location of each of said plurality of EGM signals, the method comprising, as commonly in the field, performing by a computer including one or more processors the following steps for at least one acquired mapping point, termed reference mapping point:

a) detecting each beat present in at least one recorded ECG signal and identifying a beat of interest from the detected beats;

b) identifying a principal EGM wave related to the identified beat of interest; and c) identifying an onset and an end time landmarks of said principal EGM wave providing a primary delineated EGM signal and measuring a voltage amplitude of the primary delineated EGM signal.

Contrary to the known proposals in the field, the proposed method further comprises:

d) performing a further analysis (qualitative analysis) of the primary delineated EGM signal by means of at least:

d1 ) discriminating the primary delineated EGM signal based on a decision policy that takes into account EGM signal features thereof; d2) based on said discriminating, tagging the reference mapping point with a first tag indicative that the primary delineated EGM signal becomes a double component EGM signal or with a second tag indicative that the primary delineated EGM signal becomes a single component EGM signal; and

e) creating a conducting channel map and a propagation map of the heart based on the result of said tagging.

According to a preferred embodiment, the plurality of mapping points have been acquired during a normal sinus rhythm of the patient and later stored in the signal acquisition unit. Alternatively, the plurality of mapping points can be acquired during arrhythmia rhythm of the patient. Preferably, the EGM signal features in step d1 ) are the width and amplitude of the signal. However, other EGM signal features such as the frequency, power, root mean square (RMS) value, among other signal features, could be similarly used by the proposed method to discriminate the primary delineated EGM signal.

According to an embodiment, step d1 ) comprises classifying the primary delineated EGM signal of said reference mapping point into two categories/classes, a first category indicating that the primary delineated EGM signal of said reference mapping point is a normal EGM signal and a second category indicating that the primary delineated EGM signal of the reference mapping point is an abnormal EGM signal.

Besides, according to another embodiment, the second category comprises a first sub- category/sub-class indicating that the primary delineated EGM signal of the reference mapping point is a short duration abnormal amplitude EGM signal candidate of having a second component EGM signal and a second sub-category indicating that the primary delineated EGM signal of the reference mapping point is a wide duration EGM signal candidate of having a second component EGM signal. In case of the reference mapping point is classified onto the first category, then, the reference mapping point is tagged with the second tag.

According to an embodiment, the method performs, before said step d2), and based on said discriminating step of step d1 ), a searching procedure over the primary delineated EGM signal determining existence or not of a second EGM component thereof.

The searching procedure comprises, if the primary delineated EGM signal of said reference mapping point being classified in said second category, looking for the existence of a second EGM component within a first defined searching region representing a time window corresponding to the occurrence of a QRS complex of the beat of interest in the recorded ECG signal of said reference mapping point; and/or a second defined searching region representing a time window corresponding to the rest of the cardiac cycle until the next QRS complex in the recorded ECG signal of said reference mapping point.

Moreover, if the searching procedure finds a second EGM component in the primary delineated EGM signal of said reference mapping point, the method further comprises identifying a principal wave of the second EGM component found, identifying an onset and an end time landmarks of said principal wave of the second EGM component found and measuring the voltage thereof, and tagging the reference mapping point with the first tag. Alternatively, if the searching procedure does not find a second EGM component in the primary delineated EGM signal of said reference mapping point, the reference mapping point is tagged with the second tag. If in step d2) the reference mapping point has been tagged with the first tag, the method further comprises measuring a time distance between the two EGM components of the delineated EGM signal, and based on said measured time distance, tagging the reference mapping point with a first sub-tag of the first tag indicative of a slow conducting channel entrance signal or with a second sub-tag of the first tag indicative of a slow conducting channel signal, based on a time threshold decision.

According to an embodiment, for the case of the reference mapping point being tagged with the first tag, the conducting channel map and the propagation map are created by, defining a mapping window and checking that the location of the second EGM component falls within the defined mapping window, wherein if the second EGM component falls within the defined mapping window, the voltage of the second EGM component is associated to the corresponding reference mapping point anatomical location on a conducting channel map and a time of activation of the second EGM component is associated to the corresponding reference mapping point anatomical location on a propagation map, or wherein if the second EGM component doesn't fall within the defined mapping window, voltage and time properties of the reference mapping point are being evaluated as said reference mapping point tagged with the second tag, re-tagging the reference mapping point with the second tag.

For the case of the reference mapping point being tagged (or re-tagged) with the second tag, the conducting channel map and the propagation map are created by checking the voltage of the first EGM component based on a voltage threshold value, wherein if the voltage is above said voltage threshold value, the voltage is associated to the corresponding reference mapping point anatomical location on a conducting channel map and a time of activation of the first EGM component is associated to the propagation map, or wherein if the voltage is below said voltage threshold value, a zero voltage value is associated to the corresponding reference mapping point anatomical location on a conducting channel map and the reference mapping point is removed from being mapped into the propagation map. Alternatively to this embodiment, the conducting channel map and the propagation map may be created by: for the case of the reference mapping point being tagged with the first tag, the bipolar voltage of the second EGM component is associated to the corresponding reference mapping point anatomical location on a conducting channel map and a time of activation of the second EGM component is associated to the corresponding reference mapping point anatomical location on a propagation map; and for the case of the reference mapping point being tagged or re-tagged with the second tag, the bipolar voltage is associated to the corresponding reference mapping point anatomical location on a conducting channel map and a time of activation of the first EGM component is associated to the propagation map. According to an embodiment, prior to step b), a conditioning step/strategy of the EGM signal associated to the recorded ECG signal is performed by:

selecting a time window centred on the QRS complex of the beat of interest of the recorded ECG signal and centring said selected time window on each of the rest QRS complex of the recorded ECG signal;

performing a first checking on said selected time window of whether other detected beats of the recorded ECG signal are similar and synchronized with the identified beat of interest using a shape and synchronization comparison measurement criterion;

based on the result of said first checking, synchronizing with the identified beat of interest and respect to the selected time window the other detected beats having similarities with the identified beat of interest;

performing a second checking on said selected time window, which now is synchronized with the QRS complex of the beat of interest, of whether the EGM signals associated to the other similar detected beats are similar and synchronized with the EGM signal of the beat of interest using a shape and synchronization comparison measurement criterion; and

based on the result of said second checking, synchronizing said EGM signals having similarities with the EGM signal of the beat of interest, and generating a combined EGM signal, said combined EGM signal substituting the EGM signal of the beat of interest within the selected time window.

According to another embodiment, prior to step e), and for the case of the reference mapping point being tagged with the first tag, the method further comprises searching for a plurality of other acquired mapping points tagged with the first tag and located in a neighbourhood of said reference mapping point, and for each neighbour mapping point:

i. performing a first checking of whether the identified beat of interest of the stored ECG signal of one neighbour mapping point is similar and synchronized with the identified beat of interest of the stored ECG signal of the reference mapping point using a shape and synchronization comparison measurement criterion; ii. based on the result of said first checking, if the similarity and synchronization has been proven, aligning the beat of interest of the stored ECG signal of said one neighbour mapping point with the identified beat of interest of the stored ECG signal of the reference mapping point and continue to step iii; or if the similarity and synchronization has not been proven, going back to step i, performing the first checking for another neighbour mapping point;

iii. performing a second checking of whether the second EGM component of the EGM signal of said one neighbour mapping point is similar and synchronized with the second EGM component of the EGM signal of the reference mapping point using a shape and synchronization comparison measurement criterion; and

iv. based on the result of said second checking, if the similarity and synchronization between the second EGM components has been proven, evaluating the voltage of both second EGM components using a voltage threshold value and re-tagging the mapping points depending on the result of said evaluation, and going back to step i, performing the first checking for another neighbour mapping point; or if the similarity and synchronization between the second EGM components has not been proven, going back to step i, performing the first checking for another neighbour mapping point.

According to the proposed method, step d) can be executed on an area of interest of the patient, said area of interest being predefined by using a radiographic imaging technique.

Other embodiments of the invention that are disclosed herein also include software programs to perform the method embodiment steps and operations summarized above and disclosed in detail below. More particularly, a computer program product is one embodiment that has a computer-readable medium including computer program instructions encoded thereon that when executed on at least one processor in a computer system causes the processor to perform the operations indicated herein as embodiments of the invention. Present invention could be included in current electroanatomical mapping systems such as CARTO^®, NavX™ or Rythmia™ as an update package.

Brief Description of the Drawings

The previous and other advantages and features will be more fully understood from the following detailed description of embodiments, with reference to the attached figures, which must be considered in an illustrative and non-limiting manner, in which:

Fig. 1 is a block diagram of the processing chain that can be executed by the proposed method to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue according to an embodiment of the present invention. Fig. 2 is a block diagram of the whole processing chain that can be executed by the proposed method to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue according to an embodiment of the present invention.

Fig. 3 is a block diagram of the processing chain of the EGM signal conditioning strategy proposed in an embodiment of the present invention.

Fig. 4 is a general block diagram of the processing chain of the qualitative EGM signal analysis block.

Fig. 5 is an example of the proposed decision tree for EGM signal classification.

Fig. 6 is a block diagram of the processing chain of the block diagram for the searching process of the second EGM component signal.

Fig. 7 is a block diagram of the processing chain of the tagging process of the EGM signal.

Fig. 8 is a block diagram of the processing chain of the spatiotemporal filtering strategy proposed in an embodiment of the present invention.

Fig. 9 is a block diagram of the processing chain of the maps creation block.

Detailed Description of Preferred Embodiments

Fig. 1 represents a general schematic, according to a first embodiment, of different processing blocks of the invention to execute the proposed computer implemented method. The invention is annexed to a signal acquisition unit (101 ) performed by any electroanatomical mapping system software which provides general information about the mapping points. According to the invention, the mapping points may be acquired from a patient either during a normal sinus rhythm or during arrhythmia rhythm and are further stored in the signal acquisition unit. The acquired mapping points include a plurality of ECG signals, a plurality of corresponding EGM signals, and the 3D location of each of said plurality of EGM signals. According to this first embodiment, once the mapping points are stored in the signal acquisition unit, the proposed method, for one or more acquired mapping points, detects each beat present in at least one recorded ECG signal (the surface ECG within the recording excerpt) either during the normal sinus rhythm or during the arrhythmia rhythm of the patient and identifies the beat of interest from the detected beats. For simplicity of the following description from now on the term reference mapping point RMP will be used, meaning a current mapping point the method is considering, however the following explanations likewise applies to all the acquired mapping points which are stored in the signal acquisition unit and that can be equally considered for implementing the proposed method. Once the beat of interest have been identified, the method executes an electrogram detection and delineation strategy (102) that provides detection and delineation of the bipolar EGM signal associated to the beat of interest. That is, a principal EGM wave related to the identified beat of interest is identified, and then, an onset and an end time landmarks of said principal EGM wave are also identified providing a primary delineated EGM signal. Finally, the voltage amplitude of the primary delineated EGM signal is measured.

Then, once the electrogram detection and delineation strategy has been executed, a further analysis, or qualitative analysis, of the primary delineated EGM signal associated to the beat of interest is performed (103) in order to determinate the existence or not of a second EGM component (potential) in the primary delineated EGM signal. To that end, the primary delineated EGM signal is discriminated based on a decision policy that takes into account EGM signal features thereof, preferably the width and voltage amplitude of the signal, however other signal features could be also used without departing from the scope of protection of present invention. Then, based on said discriminating step, the reference mapping point RMP is tagged with a first tag indicative that the primary delineated EGM signal becomes a double component EGM signal or alternatively with a second tag indicative that the primary delineated EGM signal becomes a single component EGM signal.

Based on the result of said tagging, the method then creates (104) a conducting channel map and a propagation map of the heart of the patient. According to a second embodiment, in this case not illustrated, the qualitative analysis of the primary delineated EGM signal also comprises performing, before the execution of said tagging, and based on the result of the discriminating step, a searching procedure over the primary delineated EGM signal to determine existence or not of a second EGM component (potential) thereof. With reference now to Fig. 2, therein it is illustrated a third embodiment of the processing blocks that can be executed by the proposed method to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue. Different to the above-described embodiments, in this case the method besides the previous described processing steps also comprises the execution of an EGM signal conditioning strategy (202) and the execution of a spatiotemporal filtering strategy of EGM signals and tags (205). The EGM signal conditioning strategy (202) is executed to include the information of the EGM signals from multiple beats in order to enhance the repetitive information included among beats of the same recording (if present). The spatiotemporal filtering strategy (205) uses information enclosed in the neighbouring mapping points of the reference mapping point RMP in order to use and remove redundant and non- useful information. It should be noted that even though in the embodiment of Fig. 2 to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue the method further includes the execution of two additional strategies, a conditioning strategy and a spatial-temporal filtering strategy, in alternative embodiments of the invention, in this case not illustrated either, a single strategy of those two strategies can be executed, that is, only the conditioning strategy or the spatiotemporal filtering strategy is implemented. Moreover, even though in the embodiment of Fig. 2 the qualitative analysis includes the execution of the searching procedure, this searching procedure is optional, so in the other embodiments where the conditioning strategy and/or the spatiotemporal filtering strategy are/is applied the searching procedure is not mandatory. Following, a detailed description of each of the processing blocks that can be implemented by the proposed method will be detailed.

Signal acquisition (101 , 201 )

The standard basic functions required to be performed by the signal acquisition unit are, as said before: I) detect each beat present in at least one recorded ECG signal (this detection provides the time landmark of the onset and end of each QRS complex and the time landmark of the R wave) and II) identify the beat of interest from the detected beats. Moreover, the signal acquisition unit also provides III) 3D location coordinates of each acquired mapping point and IV) generates a 3D mesh representing the mapping cardiac chamber where the mapping catheter moves and senses electrical activity. EGM Signal conditioning (202)

In this block the repetitive beat-to-beat information (when available) contained at each recorded ECG signal is combined in order to improve the signal quality of the EGM signal related to the beat of interest. This strategy assumes, from the signal acquisition point of view, a stable contact of the mapping catheter with the myocardium of at least two seconds in order to be able to capture more than one beat in the recording excerpt.

Fig. 3 illustrates an embodiment of the decision flow of the signal conditioning processing steps:

3.1 . Compare all present ECG beats with the beat of interest using, but not limited to, a shape and synchronization comparison measurement criterion, like the normalized cross- covariance, among others, such as the dynamic time warping or the minimum RMSE difference. For notation, x_; [n] represent the fi beat signal contained in the recording excerpt with length L and x_R [n] represent the beat of interest. L represent the length of a window centered at the QRS complex R wave of each ECG beat, either x_; [n] or x_R [n]. Therefore the normalized cross-covariance between each x{n] and x_R[n] is defined as: ∑n=o(^xR [n]-x_R)(x j [n-m] -*,)

Cj.R [m]

where ½ and x_} represent the mean value of x_R [n] and x₇ [n] , respectively and m stands for the lag between those signals. The lag value m that maximizes C_{j R} [m] (denoted as τ₇) and this maximum value (denoted as 5₇) represent the synchronization and degree of similarity, respectively, between beats x₇ [n] and x_R [n].

Only those beat x₇ [n] whose similarity value is above a certain level and also it is highly synchronized (small τ_; ) with the reference beat x_R [n], are considered similar. Whether no beat x₇ [n] is considered similar, the process stops.

3.2. If the process continues, use the value of synchronization τ₇ of all those beats x₇ [n] whose are considered similar, in order to perfectly synchronize them with the reference beat x_R [n].

3.3. Then, compare the associated EGM signal s₇ [n] from each similar beat x₇ [n] with the associate EGM signal s_R [n] from the beat of interest x_R [n] using, but not limited to, a shape and synchronization comparison measurement criterion, like the normalized cross- covariance, among others, such as the dynamic time warping or the minimum RMSE difference. This comparison can be performed, but not limited to, a similar manner to step 3.1 yielding in a degree of similarity value 5 ^GM and a synchronization value rf^GM . 3.4. If any EGM signal s_; [n] from those similar beats x₇ [n] was considered similar to the reference beat x_R [n] associated EGM signal s_R [n] and also highly synchronized (small T_; ^EGM), a combination signal ¾ [n] is obtained and substitutes the associated EGM signal s_R [n] of the reference ECG beat x_R [n] for further processing. This combination can be made using, but not limited to, the median value of all synchronized EGM signals including the EGM signal of the beat of interest. Then each s_; [n] is synchronized using rf^GM and the combination EGM signal ¾ [n] associated to the ECG beat of interest is obtained as:

sfiW = median ^ s^n— r ^GM , ... , Sj[n— T ^G ], s_fi [n]J j.

Electrogram detection and delineation (102, 203) This block represents any signal processing strategy that provides detection and delineation of the bipolar EGM signal associated to the beat of interest. The standard basic functions required to be performed by this detection and delineation strategy preferably are:

V) Provide identification of the principal EGM wave related to the beat of interest.

VI) Provide the delineation (onset and end time landmarks) of this principal EGM wave.

VII) Measure the bipolar voltage of the delineated EGM.

Qualitative EGM analysis (103. 204)

This block describes the process to qualitatively analyze the EGM signal in order to decide whether to perform a searching process that looks for the existence of a second EGM component; and based on this decision and detection, performs tagging/identification of the reference mapping point RMP.

An embodiment of this process is illustrated in Fig. 4 where it is divided in tree main sub-blocks that operate in cascade and are described in next subsections.

Qualitative EGM signal discrimination/classification: In this sub-process, see Fig. 5, it is decided, based on a decision policy that takes into account qualitative features of the EGM signal, when it is necessary to perform a searching process looking for the existence of a second EGM component. This sub-process performs a discrimination/classification of the signals based on any number of features obtained from the EGM signal like a decision tree which preferably uses the width and amplitude of the signal in order to classify the mapping point RMP.

According to an embodiment, the mapping point RMP is classified into one category of two categories/classes: a first category (or normal category, i.e. short duration but bipolar voltage ≥ threshold value), and a second category (or abnormal category, i.e. if the mapping point RMP doesn't fall within the first category and thus meaning that the mapping point is a double EGM signal candidate).

Moreover, the second category may be divided into two different sub-categories/sub-classes, a first sub-category indicating that the mapping point RMP is of short duration and abnormal amplitude (i.e. bipolar voltage < threshold value), and a second sub-category indicating that the mapping point RMP is of wide duration. For instance, according to an exemplary embodiment:

- If an EGM is considered short (e.g. width≤ 60-70 ms) and has normal amplitude (e.g. bipolar voltage ≥ 1 .5-3.5 mV), then the EGM is identified as normal/single EGM signal. - If a short EGM (e.g. width≤ 60-70 ms) has abnormal amplitude (e.g. bipolar voltage < 1 .5-3.5 mV), then the EGM is considered as candidate of having a second EGM component and need further processing.

- If an EGM is considered wide (width > 60-70 ms) directly is considered as candidate of having a second EGM component and need further processing.

It should be noted that these predefined thresholds are merely used as possible examples, being other similar values also usable in the qualitative EGM signal discrimination sub- process.

Searching process: In this sub-process, see Fig. 6, it is checked and determined the existence of a second component EGM. This searching process preferably applies over those EGM signals which are defined as abnormal EGM signals candidates of having a second EGM component.

According to a preferred embodiment, the searching process defines two searching regions: Inside/first region and outside/second region:

- Inside region: Represents a time window corresponding to occurrence of the QRS complex of the beat of interest of the recorded ECG signal. It can be defined as the union of two time windows: 1 ) spanning from the onset to the end of the QRS complex of the beat of interest (identified by general function I)) and 2) spanning from the onset to the end of the bipolar EGM signal related to the beat of interest (identified by general function VI)).

- Outside region: Represents a time window corresponding to the rest of the cardiac cycle until the next QRS complex in the recorded ECG signal. It can be defined as a time window spanning from the end of the QRS complex of the beat of interest to the onset of the QRS complex of the next beat (identified by general function I). The searching process preferably looks for a sequentially or recursively strategy, within any combination of the searching regions for signal characteristics of the EGM signal, not limited to a particular processing or transformation of the signal, that reveals the existence of a second EGM component. The identification of these signal characteristics may imply the usage of thresholds over the EGM signal of the mapping point RMP that spans within the analysis windows.

If the searching process does not identify the presence of a second EGM component, then it is identified as normal/single EGM signal. However, if the searching process finds a second EGM component, then identifies it as a double component EGM. This searching process may provide:

- The principal wave of the second EGM component.

- The onset and end time landmarks of the second EGM component principal wave. - Use function VII) to obtain the bipolar voltage of the second EGM component.

EGM signal tagging:

In this sub-process, see Fig. 7, tags are associated to the mapping point RMP based on the outcomes of the previous described sub-processes. The tags are associated depending on the qualities of each EGM signal, being those preferably the following:

- Normal EGM: This tag (or second tag) is associated to the EGM signals identified as normal/single EGM (i.e. single component EGM).

- Slow conducting channel signal: This sub-tag of the first tag (or second sub-tag) is associated to the identified double component EGM signals in which a distance metric between the first and second EGM components is above a predefined time threshold. This distance metric can be defined as, but not limited to, the time distance between the principal waves of the first EGM component and the second EGM component.

- Slow conducting channel entrance: This sub-tag of the first tag (or first sub-tag) is associated to the identified double component EGM signals in which a distance metric between the first and second EGM components is below a predefined time threshold. This distance metric can be defined as, but not limited to, the time distance between the principal waves of the first EGM component and the second EGM component.

Spatiotemporal filtering (205) This block uses spatial information of the acquired mapping points in order to remove redundant mapping points. The function of this block only applies over the mapping points tagged as slow conducting channel (either entrance or signal), that is, tagged with the first tag.

Fig. 8 illustrates an embodiment of the processing steps executed in this block:

8.1 . Search for the already acquired k mapping points tagged as slow conducting channel located in a neighborhood of the reference mapping point RMP, for instance using the

Euclidean distance metric.

8.2. For each klh neighbor mapping point, performing a first checking that compares the beat of interest of the stored ECG signal (the surface ECG beat of interest) between those mapping points using a shape and synchronization comparison measurement criterion, like the normalized cross-covariance, previously explained, among others, such as the dynamic time warping or the minimum RMSE difference.

8.3. If the surface ECG of the Mh mapping point is similar and it is synchronized to the mapping point RMP, then align them to perfect synchronization. If this do not happened, the process goes to 8.2.

8.4. After alignment, performing a second checking that compares the second EGM component of the Mh neighbor mapping point with the mapping point RMP using a shape and synchronization comparison measurement criterion, like the normalized cross- covariance, previously explained, among others, such as the dynamic time warping or the minimum RMSE difference.

8.5. If those second EGM components are not similar or do not occur synchronously, go to step 8.2. On the contrary, if those second EGM components are similar and occur synchronously, they are evaluated. 8.6. Evaluation of similar and synchronized second EGM components can be done by evaluating its bipolar voltage based on a voltage threshold value; hence, the mapping point with higher bipolar voltage retains its tag whereas the other mapping point is re-tagged to Normal/single EGM mapping point (i.e. with the second tag) depending on the result of said evaluation. Then the process goes back to step 8.2. Creation of conducting channel and propagation maps (104, 206)

This block selects the values of each EGM signal to be represented within the anatomical reconstruction of the heart. Two types of electroanatomical maps are preferably generated: the conducting channel map and the propagation map. Voltage and timing values of the EGM signal are associated to the anatomical 3D representation in conjunction with the tags. Fig. 9 illustrates an embodiment in which mapping rules are defined for each tag:

• Normal/single EGM tag (second tag): If the bipolar voltage of the mapping point RMP is above 1 .5 mV (not limitative as other thresholds are also possible), then the this bipolar voltage is associated to the conducting channel map, and the mapped propagation time associated to the propagation map, is, for example, the difference between the onset landmark to the electrical reference landmark. If the bipolar voltage is below 1 .5 mV then the voltage of 0 mV is associated to the conducting channel map and said mapping point RMP is not used for propagation map.

• Slow conducting channel entrance/signal (first tag): This process preferably defines a mapping window comprised between the 5th percentile of the second EGM component onset landmarks and the 95^th percentile of the second EGM component offset landmarks. This mapping window is used to decide whether to map the EGM qualities following, for example, the following rules:

o If the second EGM component is comprised within this mapping window, then the bipolar voltage of the second EGM component is associated to the conducting channel map and the mapped propagation time, is, for example, the time difference between the onset landmark to the electrical reference landmark, o If the second EGM component is located outside the mapping window, then it is re-tagged to Normal/single EGM (second tag) and treated as Normal/single EGM. As alternative to previous embodiment, the creation of conducting channel map and propagation map can be done following the rules:

• Normal/single EGM tag (second tag): The bipolar voltage of the first EGM component is associated to the conducting channel map, and the mapped propagation time on the propagation map, is, for example, the difference between the onset landmark to the electrical reference landmark.

• Slow conducting channel entrance/signal (first tag): The bipolar voltage of the second EGM component is associated to the conducting channel map, and the mapped propagation time on the propagation map, is, for example, the difference between the onset landmark of the second EGM component to the electrical reference landmark. The proposed invention may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions may be stored on or encoded as one or more instructions or code on a computer-readable medium.

Computer-readable media includes computer storage media. Storage media may be any available media that can be accessed by a computer. By way of example, and not limitation, such computer-readable media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Disk and disc, as used herein, includes compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media. Any processor and the storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In the alternative, the processor and the storage medium may reside as discrete components in a user terminal. As used herein, computer program products comprising computer-readable media including all forms of computer-readable medium except, to the extent that such media is deemed to be nonstatutory, transitory propagating signals.

The scope of the present invention is defined in the following set of claims.

Claims

1 . A computer implemented method to identify the ventricular arrhythmogenic substrate in myocardial scar or fibrotic tissue, wherein a plurality of mapping points acquired from a patient are stored in a signal acquisition unit, said plurality of mapping points including a plurality of electrocardiogram, ECG, signals, a plurality of corresponding intracardiac electrogram, EGM, signals, and a 3D location of each of said plurality of EGM signals, the method comprising performing by a computer including one or more processors the following steps for at least one acquired mapping point, termed reference mapping point (RMP):

b) identifying a principal EGM wave related to the identified beat of interest; and c) identifying an onset and an end time landmarks of said principal EGM wave providing a primary delineated EGM signal and measuring a voltage amplitude of the primary delineated EGM signal,

wherein the method being characterized in that it further comprises:

d) performing a further analysis of the primary delineated EGM signal by means of at least:

d1 ) discriminating the primary delineated EGM signal based on a decision policy that takes into account EGM signal features thereof; and

d2) based on said discriminating, tagging the reference mapping point (RMP) with a first tag indicative that the primary delineated EGM signal becomes a double component EGM signal or with a second tag indicative that the primary delineated EGM signal becomes a single component EGM signal; and

2. The computer implemented method of claim 1 , wherein the EGM signal features in step d1 ) at least include width and amplitude of the signal.

3. The computer implemented method of claim 2, wherein said step d1 ) comprises classifying the primary delineated EGM signal of said reference mapping point (RMP) into at least two categories, a first category indicating that the primary delineated EGM signal of said reference mapping point (RMP) is a normal EGM signal and a second category indicating that the primary delineated EGM signal of the reference mapping point (RMP) is an abnormal EGM signal.

4. The computer implemented method of claim 3, wherein said second category further comprises: a first sub-category indicating that the primary delineated EGM signal of said reference mapping point (RMP) is a short duration abnormal amplitude EGM signal candidate of having a second component EGM signal, and

a second sub-category indicating that the primary delineated EGM signal of said reference mapping point (RMP) is a wide duration EGM signal candidate of having a second component EGM signal.

5. The computer implemented method of claim 3 or 4, comprising performing, before said step d2), and based on said discriminating step, a searching procedure over the primary delineated EGM signal determining existence or not of a second EGM component thereof.

6. The computer implemented method of claim 5, wherein said searching procedure comprises, if the primary delineated EGM signal of said reference mapping point being classified in said second category, looking for the existence of a second EGM component within:

a first defined searching region representing a time window corresponding to occurrence of a QRS complex of the beat of interest in the recorded ECG signal of said reference mapping point (RMP); and/or

a second defined searching region representing a time window corresponding to the rest of the cardiac cycle until the next QRS complex in the recorded ECG signal of said reference mapping point (RMP).

7. The computer implemented method of claim 5 or 6, wherein:

if the searching procedure finds a second EGM component in the primary delineated EGM signal of said reference mapping point (RMP), the method further comprises:

identifying a principal wave of the second EGM component found; identifying an onset and an end time landmarks of said principal wave of the second EGM component found and measuring the voltage thereof; and

tagging the reference mapping point (RMP) with the first tag; or

if the searching procedure does not find a second EGM component in the primary delineated EGM signal of said reference mapping point (RMP), the reference mapping point (RMP) is tagged with the second tag.

8. The computer implemented method of any of previous claims, wherein if in said step d2) the reference mapping point (RMP) has been tagged with the first tag, the method further comprises:

measuring a time distance between the two EGM components of the delineated EGM signal; and based on said measured time distance, tagging the reference mapping point (RMP) with a first sub-tag of the first tag indicative of a slow conducting channel entrance signal or with a second sub-tag of the first tag indicative of a slow conducting channel signal, based on a time threshold decision.

9. The computer implemented method of any of previous claims, wherein said conducting channel map and said propagation map being created by:

for the case of the reference mapping point (RMP) being tagged with the first tag, defining a mapping window and checking that the location of the second EGM component falls within the defined mapping window,

wherein if the second EGM component falls within the defined mapping window, the bipolar voltage of the second EGM component is associated to the corresponding reference mapping point (RMP) anatomical location on a conducting channel map and a time of activation of the second EGM component is associated to the corresponding reference mapping point (RMP) anatomical location on a propagation map, or

wherein if the second EGM component doesn't fall within the defined mapping window, voltage and time properties of the reference mapping point (RMP) are being evaluated as said reference mapping point (RMP) tagged with the second tag, re-tagging the reference mapping point (RMP) with the second tag; or

for the case of the reference mapping point (RMP) being tagged or re-tagged with the second tag, checking the bipolar voltage of the first EGM component based on a voltage threshold value, wherein if the voltage is above said voltage threshold value, the bipolar voltage is associated to the corresponding reference mapping point (RMP) anatomical location on a conducting channel map and a time of activation of the first EGM component is associated to the propagation map, or

wherein if the voltage is below said voltage threshold value, a zero voltage value is associated to the corresponding reference mapping point (RMP) anatomical location on a conducting channel map and the reference mapping point (RMP) is removed from being mapped into the propagation map.

10. The computer implemented method of any of previous claims 1 to 8, wherein said conducting channel map and said propagation map being created by:

for the case of the reference mapping point (RMP) being tagged with the first tag, the bipolar voltage of the second EGM component is associated to the corresponding reference mapping point (RMP) anatomical location on a conducting channel map and a time of activation of the second EGM component is associated to the corresponding reference mapping point (RMP) anatomical location on a propagation map; and for the case of the reference mapping point (RMP) being tagged or re-tagged with the second tag, the bipolar voltage is associated to the corresponding reference mapping point (RMP) anatomical location on a conducting channel map and a time of activation of the first EGM component is associated to the propagation map.

1 1 . The computer implemented method of claim 1 , wherein prior to step b) a conditioning step of the EGM signal associated to said recorded ECG signal is performed by:

selecting a time window centred on a QRS complex of the beat of interest of the recorded ECG signal and centring said selected time window on each of the rest QRS complex of the recorded ECG signal;

performing a second checking on said selected time window of whether the EGM signals associated to the other similar detected beats are similar and synchronized with the EGM signal of the beat of interest using a shape and synchronization comparison measurement criterion; and

12. The computer implemented method of claim 8, wherein prior to step e) and for the case of the reference mapping point (RMP) being tagged with the first tag, the method further comprises searching for a plurality of other already acquired mapping points tagged with the first tag and located in a neighbourhood of said reference mapping point (RMP), and for each neighbour mapping point:

i. performing a first checking of whether the identified beat of interest of the stored ECG signal of one neighbour mapping point is similar and synchronized with the identified beat of interest of the stored ECG signal of the reference mapping point (RMP) using a shape and synchronization comparison measurement criterion;

ii. based on the result of said first checking,

if the similarity and synchronization has been proven, aligning the beat of interest of the stored ECG signal of said one neighbour mapping point with the identified beat of interest of the stored ECG signal of the reference mapping point (RMP) and continue to step iii; or if the similarity and synchronization has not been proven, going back to step i, performing the first checking for another neighbour mapping point;

iii. performing a second checking of whether the second EGM component of the EGM signal of said one neighbour mapping point is similar and synchronized with the second EGM component of the EGM signal of the reference mapping point (RMP) using a shape and synchronization comparison measurement criterion; and

iv. based on the result of said second checking,

if the similarity and synchronization between the second EGM components has been proven, evaluating the voltage of both second EGM components using a voltage threshold value and re-tagging the mapping points depending on the result of said evaluation, going back to step i, performing the first checking for another neighbour mapping point; or

if the similarity and synchronization between the second EGM components has not been proven, going back to step i, performing the first checking for another neighbour mapping point.

13. The computer implemented method of claim 1 , wherein the plurality of mapping points stored in the signal acquisition unit being acquired during a normal sinus rhythm of the patient.

14. The computer implemented method of claim 1 , wherein the plurality of mapping points stored in the signal acquisition unit being acquired during arrhythmia rhythm of the patient.

15. A computer program product comprising computer readable instructions that when executed in a computer system implement the steps of the method of any of the claims 1 to 14.