SYSTEM FOR CHARACTERIZING A HEART RHYTHM AND ASSOCIATED METHOD

Abstract

HEART RHYTHM CHARACTERIZATION SYSTEM AND ASSOCIATED METHOD A heart rhythm characterization system configured to implement, by computer, the following steps: for each arrhythmia in a plurality of arrhythmias, generate, using a set of elementary classifiers, indicators of the probability of the arrhythmia's presence over a time window using descriptors from a set of at least one derivation of an electrocardiogram acquired over the time window; generate an input vector (Vk) from the sets of indicators generated for the arrhythmias, the components of which include only each indicator from a selection of indicators; generate, using a global classifier (GC), global indicators of the probability of the arrhythmias' presence in the time window from the input vector (Vk).The system includes a selector (SE) allowing the selection and / or deselection of at least one of the indicators so that the indicator selection is composed of each selected indicator. Figure for the abbreviation: Fig. 3,

Description

Title of the invention: SYSTEM FOR CHARACTERIZING A HEART RHYTHM AND ASSOCIATED METHOD Field of the invention

[0001] The field of the invention is that of characterizing heart rhythms by detecting them in recordings of a biological signal. Detecting heart rhythm abnormalities allows a specialist to implement potential treatment either to relieve symptoms or to prevent more serious, potentially lethal abnormalities.

[0002] The characterization of a patient's heart rhythm is classically performed using an electrocardiogram (ECG). An ECG is the recording of the electrical activity of cardiac cells on the body surface or subcutaneously. An ECG comprises one or more views (or "leads") of this electrical activity acquired simultaneously by respective cutaneous or subcutaneous electrodes.

[0003] Electrocardiograms classically comprise 12 leads acquired over a period of 10 seconds.

[0004] The series of electrical waves of a heartbeat are conventionally designated, in their temporal order, P, Q, R, S, and T. P waves correspond to the electrical activity of the atria, and QRS waves to the electrical activity of the ventricles. By convention, the RR interval is used to designate the time between two consecutive heartbeats. In this patent application, a heartbeat is understood to be a portion of an electrical signal comprising the R wave, for example, centered on the R wave, the duration of which is the RR interval.

[0005] The use of miniature, portable devices for continuous electrocardiogram recording, such as Holter monitors, is increasing rapidly, particularly due to the aging population. Furthermore, these recordings are being made over increasingly longer periods, from 24 hours to one month, as their accuracy is directly related to the recording duration. The usual analysis, using dedicated software and subsequently supervised by a specialist, becomes impractical. For example, a 30-day recording contains 3,000,000 heartbeats. A 1% error rate in 3,000,000 beats forces the specialist to review a large number of false positives and, even more critically, risks missing false negatives.

[0006] There is therefore a growing need for optimal characterization of heart rhythms to help rhythmologists detect arrhythmias. State of the art

[0007] Automatic systems for characterizing heart rhythms are known, which process previously digitized ECGs.

[0008] One drawback of known automatic systems lies in the insufficient reliability of the results obtained, given the mass of data collected.

[0009] One object of the invention is to limit this drawback. Summary of the invention

[0010] To this end, the invention relates to a system for characterizing a heart rhythm comprising a data processing unit configured to implement, by computer, the following steps: • For each arrhythmia in a plurality of arrhythmias, generate, using a set of elementary classifiers, a set of probability indicators for the presence of the arrhythmia over a time window, using descriptors from a set of at least one derivation of a patient's electrocardiogram acquired over the time window, at least one indicator being capable of being alternately in a selected and deselected state, • generate an input vector from the indicator sets generated for arrhythmias, the components of the input vector comprising, among the indicators in the indicator sets generated for arrhythmias, only each indicator from a selection of indicators, • generate, using a global classifier, global indicators of the probability of presence of arrhythmias in the time window from the input vector.

[0011] The system includes a selector allowing the selection and / or de-selection of at least one of the indicators so that the selection of indicators is composed of each selected indicator.

[0012] Advantageously, the value of each component of the input vector associated with one of the global indicators being fixed to the same predetermined value.

[0013] Advantageously, the selector allows each indicator of a subset of at least one of the generated indicators to be selected and / or deselected individually using the elementary classifiers of the sets of elementary classifiers.

[0014] Advantageously, the set of at least one derivation comprises several derivations and the selector comprises a first selector allowing the derivations to be individually deselected and / or selected, so that each first generated indicator, using each respective elementary classifier of a sub- a predetermined set of elementary classifiers, from a selected derivation is deselected.

[0015] Advantageously, the subset consists of each elementary classifier configured to generate first indicators of the probability of presence of arrhythmias from first descriptors of the set of at least one derivation, each first indicator being generated from a single derivation.

[0016] Advantageously, the selector includes a second selector allowing for the individual selection and / or deselection of groups of elementary classifiers from sets of elementary classifiers such that each indicator generated using a selected group is in the deselected state, the groups of elementary classifiers using respective descriptors distinct from the set by at least one derivation, the elementary classifiers of the same group using the same descriptors.

[0017] Advantageously, the selector includes a third selector allowing the elementary classifiers of the classifier sets to be selected and / or deselected individually so that each indicator generated using a selected elementary classifier is in the deselected state.

[0018] Advantageously, the system includes a user interface allowing a user to select and / or deselect at least one of the elementary indicators.

[0019] Advantageously, the selector includes an automatic selector configured to: • check if a selection or deselection condition for at least one indicator is met, • automatically select or respectively deselect said at least one indicator when the selection or respectively deselection condition is met.

[0020] In a particular embodiment, the generation of the set of indicators for the probability of the presence of arrhythmia comprises: • generate an initial set of indicators for the probability of arrhythmia occurring within the time window using a first classifier that uses values ​​from at least one derivation, • generate a second set of indicators of the probability of presence of the arrhythmia in the time window, using a second classifier from first portions, preceding the R wave, of combinations of second portions of beats from the set of at least one lead, • generate a third set of indicators of the probability of the presence of the arrhythmia in the time window using a third classifier using a set of statistical indicators representative of the R_R interval distribution of the electrocardiogram.

[0021] Advantageously, the first portions are devoid of R wave. Advantageously, the data processing unit is configured to define a sequence of states of the individual's heart rhythm, taken from arrhythmias and sinus rhythm, from sets of global indicators generated for successive time windows.

[0022] Advantageously, the definition of the state sequence uses a Viterbi algorithm configured to determine the most probable state sequence likely to be obtained by a hidden Markov model from the global indicator sets.

[0023] Advantageously, the system includes an acquisition system comprising electrodes configured to acquire the entirety of at least one lead of the electrocardiogram.

[0024] Advantageously, the acquisition system is capable of forming an acquisition device comprising a processing subunit configured to implement a descriptor generator to generate the descriptors of the set of at least one derivation.

[0025] Advantageously, the arrhythmias are atrial fibrillation and atrial flutter.

[0026] The invention also relates to a method for characterizing a heart rhythm comprising the following steps, implemented by computer: • For each arrhythmia in a plurality of arrhythmias, generate, using a set of elementary classifiers, a set of probability indicators for the presence of the arrhythmia over a time window, using descriptors from at least one lead of a patient's electrocardiogram acquired over the time window, the indicators being able to be alternately in the selected and deselected states, • generate an input vector from the indicator sets generated for arrhythmias, the components of the input vector comprising, among the indicators in the indicator sets generated for arrhythmias, only each indicator from a selection of indicators, • generate, using a global classifier, global indicators of the probability of the presence of arrhythmias in the time window from the input vector,

[0027] select or deselect at least one of the indicators so that the indicator selection is composed of each selected indicator.

[0028] The invention also relates to a computer program product comprising a readable information carrier, on which is stored a computer program comprising program instructions, the computer program being loadable onto a data processing unit and adapted to drive the implementation of steps or stages of the process according to the invention, when the computer program is implemented on the data processing unit.

[0029] The invention also relates to a readable information carrier comprising program instructions forming a computer program, the computer program being loadable onto a data processing unit and adapted to trigger the implementation of one or more steps of the process according to the invention when the computer program is run on the data processing unit. Brief description of the figures

[0030] Other features and advantages of the invention will become apparent from the following detailed description, with reference to the accompanying figures, which illustrate:

[0031] [Fig-1]: a schematic view of an example of a system according to the invention,

[0032] [Fig.2]: a schematic view of a derivation,

[0033] [Fig.3]: a view of functional blocks of an example of a system according to the invention,

[0034] [Fig.4]: a representation of a heartbeat,

[0035] [Fig.5]: a block diagram of the steps of an example of the process according to the invention. Description of the invention

[0036] The invention relates to a method for characterizing a heart rhythm and to a SYS system configured to implement the method according to the invention.

[0037] The SYS system includes, for example, as shown in [Fig. 1], an ACQ acquisition system for an electrocardiogram, denoted ECGk hereafter, of a patient during an acquisition time window denoted Fk ​​hereafter. Acquisition system

[0038] The ACQ acquisition system includes a set E of cutaneous or subcutaneous electrodes allowing simultaneous acquisition during the same time window Fk of acquisition of duration T, of a set of at least one elementary analog signal Sk; analogs with i = 1 to N and N is an integer greater than or equal to 1 representing the number of derivations of the electrocardiogram ECGk acquired on the time window.

[0039] N is advantageously between 1 and 12, but can, alternatively, be greater than 12.

[0040] Each elementary signal Sk; represents the evolution of the value Vi(t) of the potential difference between electrodes of the set E as a function of time t during the duration T of the acquisition window Fk.

[0041] The ACQ acquisition system also includes an analog-to-digital converter AN to convert the elementary analog Sk signals into digital signals called derivations Dk; in the following text, the signals being sampled at a predetermined sampling frequency.

[0042] The sampling frequency is advantageously between 100 Hz and 2000 Hz, preferably between 100 Hz and 1000 Hz. It is, for example, equal to 200 Hz.

[0043] The acquisition device may include a combiner configured to combine digital signals so that at least one derivation is a combination of digital signals.

[0044] Each derivation Dkj is one of the digital signals composing the patient's electrocardiogram ECGk, acquired during the acquisition time window of duration T. It is a sequence of potential difference values ​​Vi taken at regular time intervals over the duration T. An example of a derivation Dkj is shown schematically in [Fig.2].

[0045] The acquisition system may include derivation processing modules, for example to filter the derivations from the analog converter, before delivering the Dk derivations;.

[0046] The duration T of the acquisition window is advantageously greater than or equal to 10 seconds. It is, for example, between 10 seconds and 30 minutes. It is, for example, equal to 60 seconds. It is defined so that each lead includes a plurality of heartbeats.

[0047] The electrode assembly E may include at least one electrode intended to be installed on the surface of the skin and / or at least one electrode intended to be installed subcutaneously.

[0048] Alternatively, the SYS system is devoid of the acquisition system and is intended to receive the leads of an ECGk of a patient acquired during the acquisition time window Fk by an external acquisition system, or directly descriptors of such leads as we will see later in the description. Processing system

[0049] The SYS system includes a processing system S configured to process, by computer, the set of at least one derivation Dkj or the descriptors of the set of at least one derivation Dkj so as to generate indicators of the probability of presence of arrhythmias of a predetermined set of arrhythmias, during the acquisition window Fk.

[0050] By probability of presence of an arrhythmia during the acquisition window Fk generated from a derivation, we mean the probability that the arrhythmia is present during the time window Fk

[0051] In the example of [Fig.2], the set of arrhythmias consists of atrial fibrillation, noted AF, also called atrial fibrillation, and atrial flutter, noted AFL, also called atrial flutter.

[0052] This example is not limiting. In general, the set of arrhythmias includes several arrhythmias taken from among atrial fibrillation, atrial flutter, ventricular tachycardia, supraventricular tachycardia, at least one bradycardia, for example atrioventricular block.

[0053] The system S comprises a data processing unit TR, for example a processor assembly PO comprising one or more processors, and a memory assembly MEM comprising one or more memories in which the system's functional components are stored, implemented as software or code executable by the processor assembly PO. The data processing unit is configured to implement the arrhythmia characterization process by executing the functional components stored on the memory assembly MEM, i.e., by using the various functional components described below. The steps of the arrhythmia characterization process are the steps implemented by executing the various functional components. Functional bricks

[0054] Figure 3 shows an example of a processing system S on which functional building blocks of the processing system S are shown.

[0055] The system S includes a classifier C configured to generate a set of probability indicators for the presence of the arrhythmia over the time window Fk, for each time window Fk of a succession of time windows (with 1 = 1 to K where K is the number of time windows) and for each arrhythmia in a set of AF, FLA arrhythmias. The set of probability indicators is generated from descriptors of at least one lead Dk of an electrocardiogram (ECGk) acquired during the time window Fk.

[0056] The processing system S advantageously includes a descriptor generator GD configured to generate the descriptors from the set of at least one derivation Dk;. Alternatively, the descriptor generator GD is external. Elementary classifiers

[0057] To this end, the classifier C comprises a set of elementary classifiers for each arrhythmia. In other words, in the case of M arrhythmias, the classifier C comprises M sets of elementary classifiers. M is an integer greater than or equal to 2.

[0058] Thus, in the example of [Fig.3], the classifier C comprises a first set of elementary classifiers MO_FA, RR_FA, PW_FA associated with fibrillation atrial, that is to say configured to generate a first set of indicators IMFAki, IMFAk2, IRRFAk, IPFAki and IPFAk2 representative of probabilities of presence of atrial fibrillation, and a second set of elementary classifiers MO_FLA, RR_FLA, PW_FLA associated with atrial flutter, that is to say configured to generate a second set of indicators IMFAki, IMFLAk2, IRRFLAk, IPFLAki, IPFLAk2 representative of probabilities of presence of atrial fibrillation AF.

[0059] The different elementary classifiers associated with the same arrhythmia, that is, from the same set of elementary classifiers, belong to distinct groups in the sense that they are configured to generate the respective indicators from distinct descriptors of the different Dk derivations; or of the derivation in the case of using a single derivation. Conversely, elementary classifiers associated with different arrhythmias and using the same descriptors belong to the same group.

[0060] Advantageously, each elementary classifier is configured to discriminate a derivation characterizing the presence of an arrhythmia from a derivation characterizing the presence of a sinus rhythm, from the descriptors of a derivation.

[0061] To this end, each classifier was previously trained using training lead descriptors, acquired over the same duration T and having the same sampling frequency as the leads in the set of at least one lead. The training leads comprise only initial training leads characterizing a sinus rhythm and other training leads, the other training leads characterizing the presence of the arrhythmia processed by the classifier so as to discriminate a lead of atrial fibrillation (AF) from a lead of sinus rhythm (RS).

[0062] Advantageously, each set of elementary classifiers associated with an arrhythmia includes a first classifier, called morphological, a second classifier, called P-wave analyzer and a third classifier called RR analyzer.

[0063] These classifiers are each configured to distinguish one of the arrhythmias of sinus rhythm on the basis of one of the following analyses, and (i) the morphology of atrial electrical activity, (ii) the presence or absence of a P wave of atrial activity, (iii) the presence or absence of an irregularity of the heart rhythm.

[0064] In the patent application, the indicators delivered by the different elementary classifiers are noted XYk where X corresponds to the type of classifier and Y is the arrhythmia sought, i.e. of which the classifier generates a probability of presence, k is the temporal order number of the time window Fk in the sequence of time windows.

[0065] In the example of [Fig.3], we note X= MO for the morphological classifier, X= RR for the RR analyzer classifier and X= PW for the P wave analyzer classifier and Y= FA for atrial fibrillation and Y= FLA for flutter.

[0066] Some indicators are supplemented by an index i. Indicators with index i are generated from features of a first derivation with index i, Dk, from the set of at least one derivation. For example, the indicators IMFAki and IMFLAki are generated from features of the first derivation Dkb. The indicators IMFAk2 and IMFLAk2 are generated from features of the second derivation Dk2. Morphological classifier

[0067] The so-called morphological classifier MO_FA, respectively MO_FLA, associated with a given arrhythmia FA, respectively FLA, is configured to process consecutive values ​​Vi of the derivation(s) Dki, Dk2 to generate a first set of indicators IMFAki, IMFAk2; respectively IMFLAki, IMFLAk2, representative of probabilities of presence of the arrhythmia FA, respectively FLA, during the time window Fk.

[0068] Each indicator IMFAki, IMFAk2, IMFLAki, IMFLAk2 is generated from consecutive values ​​Vi of a single derivation Dkb Dk2. Each morphological classifier MO_FA, respectively MO_FLA, is configured to generate indicators IMFAki, IMFAk2, respectively IMFLAki, IMFLAk2 from the values ​​of the derivation(s) Dkb Dk2.

[0069] Advantageously, in the case of several derivations, each morphological classifier MO_FA, MO_FLA is configured to process the consecutive values ​​of the two derivations Dki Dk2 so as to successively generate the indicators IMFAki and IMFAk2.

[0070] Advantageously, each morphological classifier MO_FA, MO_FLA is an artificial neural network, for example a convolutional neural network or CNN. Alternatively, the classifier is a transformer or a recurrent neural network.

[0071] Each morphological classifier MO_FA, respectively MO_FLA is advantageously an artificial neural network trained from consecutive values ​​of derivations acquired over the same duration and that of the acquisition window Fk with the same sampling frequency, some of which characterize a sinus rhythm and others characterize the presence of the arrhythmia FA, respectively FLA, processed by the elementary classifier so as to discriminate a derivation characterizing an atrial fibrillation FA from a derivation characterizing a sinus rhythm RS.

[0072] This type of classifier and discriminator for arrhythmias generates distinct characteristic distributions from one arrhythmia to another and from those of a patient with a sinus rhythm.

[0073] This is, for example, the case with atrial fibrillation and flutter. Indeed, atrial fibrillation is characterized by irregular electrical activity, whereas the electrical activity in the presence of a sinus rhythm is regular and is between 60 and 100 beats per minute, while the electrical activity in the presence of atrial fibrillation is between 300 and 600 cycles per minute.

[0074] Flutter, on the other hand, is characterized by regular electrical activity. The heart rate in the presence of flutter is between 100 and 300 cycles per minute. P-wave analyzer classifier

[0075] The classifier C further includes a classifier called a P-wave analyzer for each arrhythmia.

[0076] Each P-wave analyzer classifier PW_FA; respectively PW_FLA, associated with a given FA, respectively FLA arrhythmia, is configured to process combinations, for example averages, of predetermined portions of beats present on the respective leads Dki, Dk2, acquired during the acquisition window F to generate IPFAi, IPFA2; respectively IPFLAi, IPFLA2, indicators of the probability of presence of the arrhythmia.

[0077] These portions are advantageously portions preceding the R wave and devoid of the R wave.

[0078] These portions are, for example, located only before the R wave.

[0079] These portions advantageously include the P wave of depolarization temporally preceding the R wave in the case of sinus rhythm or, more generally, the part of the combination likely to include the P wave in the case of sinus rhythm.

[0080] Advantageously, the combined portions, for example averaged, correspond to portions of the same predetermined duration preceding the wave R and devoid of Fonde R.

[0081] Each portion is, for example; a portion of predetermined duration preceding the R wave. This duration is, for example, between 200 ms and 500 ms.

[0082] For example, the predetermined duration portion extends from a first instant to a second instant preceding the R wave and separated by a predetermined time gap from the previously detected R wave or the previously detected QRS complex.

[0083] Alternatively, each portion corresponds to a complete beat.

[0084] In another embodiment, the portion follows Fonde R. It is, for example, located only after Fonde R.

[0085] The system S advantageously comprises a combiner, for example an averager MO, of P-waves configured to generate, for each derivation Dki, respectively Dk2, a combination, for example an average Mkb Mk2 of at least a subset of the beats present on the derivation DkB respectively Dk2 and detected by the wave detector R. The combined beats, for example averaged, advantageously have the same duration equal to the period of the beat, are, for example, centered on the R wave and have a predetermined duration, advantageously defined so that the beats used for the combination or averaging include the successive waves P, Q, R, S and T.

[0086] Alternatively, the R waves are subtracted from the combined waves before the combination, or the R waves are subtracted from the combinations. This makes classification easier.

[0087] Combining or averaging portions that include only the P wave (or, more generally, the portion of the beat that may include the P wave) makes it easier to classify atrial activity.

[0088] The processing system S advantageously, but not necessarily, includes an EP portion extractor configured to extract a portion Pki, Pk2 from one of the combinations or averages Mkb Mk2 of the beat, obtained from one of the leads Dkb Dk2. This portion is, for example, the combination or average under consideration; it may include a portion preceding the R wave and the R wave, for example, the entire combined or averaged PQRS complex, or a portion, for example of predetermined duration, of the combined or averaged beat, preceding the R wave and lacking the R wave, that is, including only a part of the combination preceding the R wave. Suppression of the QRS complex, which is significantly more intense than the preceding portion Fonde P, particularly in cases of atrial fibrillation or flutter, improves the performance of the classifier.

[0089] This portion is, for example; a portion of predetermined duration preceding the R wave. This duration is, for example, between 200 ms and 500 ms.

[0090] For example, the predetermined duration portion extends from a first instant to a second instant preceding the R wave and separated by a predetermined time gap from the previously detected R wave or the previously detected QRS complex.

[0091] Each P-wave analyzer classifier PW_FA, PW_FLA is configured to generate IPFAkb IPFAk2 indicators; respectively IPFLAki, IPFLAk2 of probabilities of presence of the given arrhythmia from Pkb Pk2 portions of combinations or beat averages of Dkb Dk2 leads.

[0092] Each P-wave analyzer classifier PW_FA, PW_FLA is advantageously configured to generate each arrhythmia probability indicator IPFAki, IPFAk2; IPFLAki, IPFLAk2 from one of the portions Pkh Pk2. Each P-wave analyzer classifier, PW_FA, respectively PW_FLA is capable of generating indicators IMFAki, IMFAk2, respectively IMFLAki, IMFLAk2 from the portions of the combinations or averages resulting from the different derivations Dki, Dk2.

[0093] Advantageously, in the case of several derivations, each P wave analyzer classifier PW_FA, PW_FLA is configured to successively process the portions of the combinations from the different derivations DkiDk2, that is to say to process the derivations consecutively, so as to successively generate the indicators IMFAkj and IMFAk2.

[0094] Advantageously, each P-wave analyzer classifier PW-FA, PW_FLA is an artificial neural network, for example a convolutional neural network.

[0095] Each P-wave analyzer classifier PW_FA, PW_FLA is advantageously pre-trained from portions of combinations, for example of averages, beats of training leads acquired over the same duration as the acquisition window Fk and some of which characterize the sinus rhythm and others the arrhythmia associated with the classifier so that the classifier is able to discriminate a portion characteristic of a sinus rhythm from a portion characteristic of the arrhythmia.

[0096] P-wave analyzer classifiers are particularly interesting and discriminating for helping to detect arrhythmias that generate characteristic P-wave shapes distinct from one arrhythmia to another and from the characteristic P-wave shape of a healthy patient.

[0097] This is, for example, the case of atrial fibrillation and atrial flutter. Indeed, atrial fibrillation is characterized by an absence of a dome-shaped sinus P wave, replaced by irregular f waves, whereas atrial flutter is characterized by a "sawtooth" P wave, identical and regular, biphasic (of which the negative phase predominates: this is then referred to as an F wave), whereas a sinus rhythm presents a P wave.

[0098] Moreover, a classifier is more precise and therefore more sensitive and reliable than the use of a threshold to discriminate a sinus rhythm from atrial fibrillation or flutter which are characterized by the replacement of a P wave by irregular or regular waves.

[0099] Alternatively, the combination is, for example, a median of beats.

[0100] Alternatively and / or in addition, at least one set of elementary classifiers dedicated to an arrhythmia may include at least one other classifier, for example, a classifier receiving a combination or average of QRS complexes and a combination or average of P waves separately to discriminate, for example, from an atrioventricular block of sinus rhythm, with a number of P waves greater than the number of QRS, or a sequence of P waves not followed by QRS.

[0101] As an alternative and / or in addition, a classifier configured to discriminate between derivations characteristic of an arrhythmia and derivations characteristic of sinus rhythm may be provided from a Fourier transform of at least one of the aforementioned descriptors.

[0102] Regularity analyzer classifier or regularity classifier

[0103] The classifier C advantageously includes an RR interval analyzer classifier for each arrhythmia.

[0104] Each interval analyzer classifier RR, RR_FA; respectively RR_FLA, associated with a given arrhythmia FA, respectively FLA, is configured to process statistical indicators representative of the distribution of RR intervals of the ECGk acquired during the acquisition window Fk to generate indicators IRRFAk; respectively IRRFLAk, representative of probabilities of presence of the arrhythmia FA, respectively FLA during the window Fk.

[0105] The portion of a lead Dk corresponding to a heartbeat consists of a succession of electrical waves in time, the shape of which is shown in [Fig. 4]. The P wave is the first in time and corresponds to atrial depolarization; it has a low amplitude and a dome shape in sinus rhythm. Next, the PR interval reflects the atrioventricular conduction time. The QRS complex reflects ventricular contraction, and finally, the R wave represents ventricular repolarization following the QRS complex. By convention, the R peak is considered a marker of ventricular systole. Since the R wave is most often the narrowest and widest part of the QRS complex, it is generally used by convention to temporally locate the heartbeat with very good accuracy, in practice on the order of a millisecond.Thus, the time interval between two successive R waves, also conventionally called the RR interval, precisely characterizes the time separating two successive heartbeats. The RR interval is the instantaneous period of the ECG signal. The instantaneous heart rate is the inverse of the RR interval.

[0106] The RR interval is common to all Dk leads of an electrocardiogram.

[0107] The GD descriptor generator advantageously includes a generator GIS statistical indicators characterizing the distribution of the patient's RR intervals during the acquisition window Fk from at least one derivation of the set of at least one derivation Dk; The generated statistical indicators are provided to RR_FA, RR_FLA interval analyzer generators as we will see later in the text.

[0108] The GD descriptor generator includes a beat detector DB configured to detect and time-stamp, i.e. to determine the times of occurrence or TOC, acronym for the Anglo-Saxon expression, of consecutive R waves from at least one of the leads, the distribution of R waves being the same for all leads and optionally to detect and time-stamp other waves, for example, P and / or Q and / or S and / or T from the lead.

[0109] The GIS statistical indicator generator includes a GRR series RR, SRRk generator, configured to generate a series RR, SRRk for the time window Fk, from at least one derivation of the set of derivations, for example derivation Dk;.

[0110] The RR series GRR generator includes an RR series constructor configured to construct the RR series by calculating the RR intervals between consecutive beats, from the timestamps of consecutive R waves.

[0111] The series comprises H terms R-Rk(h) with h= 1 to H-1, where H is the number of heartbeats present in the derivation Dk;, h corresponds to the time order number of the first beat from which the RR interval is calculated, considered in the derivation Dki.

[0112] The RR series, SRRk comprises successive values ​​that are consecutive R-Rk(h) values ​​of the instantaneous RRh intervals of the chosen derivation Dk; during the time window Fk, or that are determined from several derivations.

[0113] Alternatively, the DB beat detector is configured to detect and time-stamp one of the other depolarization waves (P, Q, S, or T) in consecutive beats. The RR series constructor then uses these timestamps to construct the RR series. However, the accuracy of the characterization of the time interval between two successive heartbeats obtained by this method is lower than when using R waves.

[0114] The GIS statistical indicator generator includes an elementary statistical indicator generator RR_GM configured to generate, from the RR series, at least one statistical indicator Gk characterizing the statistical distribution of the intervals R-Rk(h) of the RR series, SRRk.

[0115] This elementary generator RR_GM is advantageously configured to generate, from the RR series, parameters of a Gaussian mixture model characterizing the distribution of the RR intervals of the RR series, SRRk. In other words, the elementary generator RR_GM is configured to parameterize a Gaussian mixture model of so that the Gaussian mixture model models the distribution of RR intervals of the RR series.

[0116] The Gaussian mixture model is commonly referred to by the English acronym GMM, from the English expression "Gaussian Mixture Model". The Gaussian mixture model is used to parametrically estimate the distribution of the RR intervals of the RR series by modeling it as the sum of several Gaussians. The parameterization of the model consists of determining the variance, mean, and magnitude of each of the Gaussians in the model. These parameter values ​​are optimized according to the maximum likelihood criterion in order to approximate the desired distribution as closely as possible. This optimization is, for example, performed using a well-known iterative procedure called expectation-maximization (EM).

[0117] The elementary generator RR_GM is configured to calculate at least one elementary statistical indicator from at least one parameter of the Gaussian mixture model. It is, for example, configured to calculate the ratio of the means of the Gaussians of the model.

[0118] The Gaussian model includes, for example, 1 or 2 Gaussians.

[0119] The statistical indicators derived from the parameters of the Gaussian model are characteristic and discriminating. In the case of a model comprising three Gaussians, sinus rhythm comprises three Gaussians with the same center due to the stability of the RR interval. The ratio of the Gaussian means is equal to 1. It is known that the cardiac cycle of atrial flutter can only take on values ​​that are multiples of a base duration. The ratio of the Gaussian means can therefore only take on a predetermined number of values. This indicator is thus particularly relevant for highlighting the presence of atrial flutter. The distribution of the RR intervals is therefore modeled by respective Gaussians centered on these three respective values. The distribution of RR intervals in atrial fibrillation is very irregular. The ratio of the Gaussian means can therefore take on an unlimited number of values.

[0120] In the non-limiting example of [Fig.3], the GIS statistical indicator generator includes a generator of a sequence of ratios between consecutive RR intervals given by the following formula:

[0121] RAk(h) = R-Rk(h+1) / R-Rk(h) for h = 1 to H-2.

[0122] This series of ratios is characteristic of atrial fibrillation and flutter and allows these arrhythmias to be discriminated between each other and from a sinus rhythm.

[0123] Indeed, in the case of sinus rhythm, the RA(h) ratios all have a value approximately equal to 1 insofar as the heart rate is approximately stable and even if it increases, for example during exertion, the RR interval between a first beat and the second beat is approximately equal to the RR interval between the The second and third beats. In atrial flutter, the RA(h) ratio can only take a certain number of known discrete values. In atrial fibrillation, the RAk(h) ratios can take a large number of values.

[0124] Advantageously, the GIS statistical indicator generator includes a generator of indicators of the probability of presence of arrhythmias from the sequence of ratios between RR intervals.

[0125] Advantageously, the generator includes ratio classifiers RR, RA_FA, RA_FLA. Each ratio classifier is configured to discriminate between one of the arrhythmias FA or FLA from sinus rhythm.

[0126] To this end, each of the ratio classifiers is previously trained by series of ratios RAk(h) generated from training derivations, some of which are characteristic of the arrhythmia considered and others of the sinus rhythm, in order to discriminate the series of ratios from the sinus rhythm from those from the arrhythmia considered.

[0127] Each ratio classifier is configured to generate a statistical indicator IFAk, respectively IFLAk, which is an indicator of the probability of the presence of one of the AF or FLA arrhythmias considered from the series of ratios RAk(h).

[0128] We thus obtain statistical indicators Gk, IFAk, IFLAk of ​​probability of presence of the respective arrhythmias in the considered window Fk.

[0129] Ratio classifiers are, for example, artificial neural networks. These classifiers are, for example, CNNs. Alternatively, at least one ratio classifier is a transformer or a recurrent neural network.

[0130] The statistical indicators generated by the GIS statistical indicator generator are provided to the RR, RR_FA and RR_FLA classifier analyzers.

[0131] Advantageously, each RR analyzer classifier, RR_FA, RR_FLA is an artificial neural network, for example a multilayer perceptron or MLP, an acronym for the Anglo-Saxon expression "Multilayer Perceptron".

[0132] This classifier RR_FA, RR_FLA receives as input a vector composed of at least one statistical indicator generated by the GIS generator, for example the indicator Gk characterizing the R_R distribution and the probability indicator generated for the arrhythmia considered IFAk, IFLAk.

[0133] Each classifier analyzer of R_R RR_FA, RR_FLA is advantageously pre-trained from the same vector generated by the GIS statistical indicator generator from derivations, for example, measured during the same duration as the training window and of which some characterize the arrhythmia considered and others characterize the sinus rhythm, in order to discriminate these arrhythmias.

[0134] In the realization of [Fig. 1], the ACQ acquisition system is an acquisition device, that is to say an object, connected to the system S, by means of communication of the system SYS, for example wireless or wired allowing unidirectional or bidirectional communication between the system S and the ACQ acquisition device so that the ACQ acquisition device is able to transmit the derivations to the system S. These are for example the communication means described in the rest of the text.

[0135] In a particular embodiment, the processing unit comprises a first processing subunit configured to implement the generator of GD descriptors from the set of at least one derivation Dk; or at least the GIS statistical indicator generator and, preferably, the beat detector, and a second processing subunit configured to implement the C classifiers.

[0136] Advantageously, the ACQ acquisition system is capable of forming an acquisition device comprising the electrode assembly E and the first processing subunit. It comprises, for example, the electrode assembly and a housing capable of being mechanically connected to the electrode assembly or mechanically connected to it and containing the first processing unit.

[0137] Advantageously, the housing also contains the analog-to-digital converter.

[0138] The ACQ acquisition device is then connected to the second processing subunit, by means of communication of the SYS system, for example wireless or wired allowing unidirectional or bidirectional communication between the second processing subunit and the ACQ acquisition device so that the ACQ acquisition device is able to transmit the descriptors it has generated to the second processing subunit so that it can execute the classifier C.

[0139] In material terms, each processing subunit can be realized in one of the forms described above for the processing unit.

[0140] The ACQ acquisition device is advantageously portable or implantable on an individual subcutaneously. Segmentation of the approach

[0141] The use of classifiers that analyze distinct descriptors of one or more leads in parallel to generate indicators of the probability of the presence of different arrhythmias requires less training data than a single classifier that analyzes consecutive values ​​of the lead(s) to obtain indicators of equivalent reliability or sensitivity. Indeed, the segmented analysis of the lead descriptors makes it possible to limit the number of degrees of freedom of the neural networks. The training of the neural networks is thus facilitated.

[0142] Moreover, the segmented approach is more robust to artifacts present on the derivations.

[0143] Furthermore, performing in parallel a morphological analysis of the signal, a statistical analysis of the temporal distribution of the ECG signal, and a P-wave analysis to generate indicators and global probability indicators from these probabilities is particularly effective in discriminating between atrial flutter, atrial fibrillation, and sinus rhythm. As previously discussed, the descriptors analyzed by these different classifiers (morphology, P-wave, statistical indicators of the RR interval distribution) exhibit characteristics (shapes, values) typical of the different arrhythmias and sinus rhythm, and discriminate between the different arrhythmias and sinus rhythm. These descriptors are understandable to the end user because they simulate the basic rules of clinical electroradiology.

[0144] Furthermore, carrying out different analyses makes it possible to maintain good performance when one of the classifiers is less efficient for a lead with a particular appearance, for example in the case of an abnormally regular rhythm in the case of atrial fibrillation.

[0145] The generation of different probability indicators for the same arrhythmia and the same lead, by the different classifiers, allows a rhythmologist to whom these indicators are provided, to improve his knowledge of the descriptors of the analyzed rhythm compared to the generation of a single indicator, which allows him to more easily identify an arrhythmia characterized by the ECGk or to identify an elementary classifier not relevant for the detection of an arrhythmia.

[0146] Furthermore, analyzing arrhythmias using different sets of elementary classifiers improves classification performance and limits the amount of training data.

[0147] In addition, analyzing the leads independently allows the rhythmologist to select the indicators he wants to deduce information about the rhythm, especially when a lead is obviously degraded, for example when it has been incorrectly installed on the patient, which allows him to more easily and reliably identify an arrhythmia characterized by the ECGk.

[0148] Generating time window probability indicators for each time window allows information to be provided to the electrophysiologist on a regular basis throughout the acquisition of a global ECG comprising a succession of time windows. This allows for faster detection of arrhythmias and more precise identification of the time of onset of arrhythmias. Input vector generator

[0149] The system advantageously includes a vector generator GV configured to generate an input vector V from the sets of probability indicators generated by the classifier C.

[0150] The vector V is composed of a number A of components, where A is greater than or equal to the number L of probability indicators generated by the classifier C.

[0151] The vector Vk is suitable such that each of a number L of its components is one of the probability indicators. Each probability indicator corresponds to a predetermined component, that is to say to a predetermined order number in the order of the components of the vector.

[0152] In the non-limiting example of [Fig.3], the number L of global indicators is equal to 10. 10 components of the vector V are reserved for probability indicators, in the sense that their values ​​are likely to be those of the L probability indicators.

[0153] When A is greater than L, the AL other components of the vector may include values ​​of other parameters such as age, patient history, and / or risk factors. Global Classifier

[0154] The SC classification system includes a global classifier GC configured to generate a set of global indicators IGFAk, IGFLAk, IGRSk of probabilities of presence of the arrhythmias FA, FLA, and of the sinus rhythm noted RS, on the time window Fk, from an input vector noted Vk.

[0155] The global classifier GC is pre-trained from training vectors generated from training derivation descriptors, by the classifier C, to discriminate vectors V from the different arrhythmias from each other and from the sinus rhythm.

[0156] The test leads include test leads characteristic of sinus rhythm, and sets of test leads characteristic of the respective arrhythmias, each set of test leads being characteristic of one of the arrhythmias.

[0157] The global classifier GC is, for example, an artificial neural network.

[0158] The neural network is, for example, a multilayer perceptron, a feedforward neural network. Alternatively, the global classifier is a forest of decision trees.

[0159] Generating reliable global indicators from indicators generated by different classifiers requires less training data than using a single classifier. This solution also offers better sensitivity and greater robustness. Derivation sequence

[0160] Advantageously, the ACQ acquisition system is configured to generate a sequence of sets of at least one derivation, i.e. of ECGk with k= 1 to K, acquired during K successive time windows F, for example consecutive, of an overall acquisition window FG of overall duration TG.

[0161] The overall duration TG of the overall acquisition window FG is advantageously between 20s and 30 days.

[0162] Advantageously, successive time windows have the same duration.

[0163] The classifier C is configured to generate sequences of K sets of global indicators IGFAk, IGFLAk, IGRSk from the descriptors of sets of at least one successive derivation Dki, Dk2. Each set of global indicators IGFAk, IGFLAk, IGRSk is generated from a single set of at least one derivation Dki, Dk2 of the sequence of time windows Fk. The temporal order of the sets of global indicators IGFAk, IGFLAk, IGRSk is the same as that of the derivation sets DkB Dk2 from which they are respectively generated. Time analyzer

[0164] Advantageously, the system S includes a TC time analyzer configured to define a sequence of Ek states of the heart rhythm corresponding to the Ek states of the heart rhythm on successive time windows Fk from the sequence of global indicator sets IGFAk, IGFLAk, IGRSk generated for successive time windows Fk.

[0165] Each stateEk is taken from among the arrhythmias and the sinusoidal rhythm FA, FLA or RS.

[0166] The idea is to determine, from a sequence of K sets of global indicators IGFAk, IGFLAk, IGRSk, a sequence of the Ek states of the heart rhythm in the successive time windows Fk.

[0167] The TC time analyzer is configured to generate the states Ek of the respective time windows Fk from the global indicators generated for other time windows, for example, all other time windows. The TC time analyzer makes it possible to obtain a sequence of elementary states that are consistent with each other and with the known, for example, learned, probabilities of transitioning from one state to another from one time window to another. It is known, for example, that when a rhythm is fluttering in a time window, the probability of remaining fluttering in subsequent time windows is very high. On the other hand, when the rhythm is atrial fibrillation, the patient may transition to flutter, but with a low probability.

[0168] The TC temporal analyzer uses, for example, a model such as a hidden Markov model, also called an HMM (Hidden Markov model), by implementing a Viterbi algorithm to generate the most probable state sequence from the sequence of global indicators that can be generated by the hidden Markov model.

[0169] The hidden Markov model used is parameterized.

[0170] In order to parameterize the hidden Markov model, the latter was previously subjected to a first training phase during which the hidden Markov model is trained from sequences of sets of global indicators IGFAk, IGFLAk, IGRSk labeled by a sequence of known states so as to generate a matrix of transition probabilities between possible states.

[0171] The hidden Markov model was also trained, in a second training phase, to learn, for each possible state (FA, FLA, and RS), which observation (i.e., which set of global indicators IGFAk, IGFLAk, IGRSk) it corresponds to. This training phase is performed using global training indicators generated by the global classifier from descriptors of one or more derivations characteristic of the given state. The second training phase yields a matrix of observations. Separating these two training phases makes it a simple learning process.

[0172] The probability of starting in each state is, for example, predetermined. For instance, one might decide that the states are equally probable or that the probabilities of starting in the respective states are the same. Alternatively, these probabilities are calculated from a set of derivation sequences whose state sequences, and therefore the first states, are known.

[0173] Alternatively, the temporal analyzer TC includes, for example, a transformer or self-attentive model. Alternatively, the temporal analyzer TC includes a recurrent neural network, also called an RNN (Recurrent Neural Network), for example, a recurrent network with long-term memory (LSTM). Selector

[0174] In general, the processing unit is configured to generate, for each arrhythmia in a plurality of AF, FLA arrhythmias, a set of indicators of the probability of the presence of the AF, FLA arrhythmia over a time window Fk, using descriptors from a set of at least one derivation of a patient's ECGk acquired over the time window Fk. The set of elementary classifiers includes, for example, at least one of the elementary classifiers from [Fig. 2] and, optionally, at least one other elementary classifier from the S system.

[0175] It is important to note that the indicators generated by the elementary classifiers are not always relevant, their consideration by the global classifier leading to a degradation of the reliability of the global indicators.

[0176] This is, for example, the case when atrial fibrillation is coupled with a conduction disorder of the electrical impulse that modulates the RR interval. Analysis of the RR interval is not relevant for characterizing atrial fibrillation in this case, as taking into account the indicators generated by the RR analyzer classifier degrades the reliability of the indicators generated by the global classifier. The reliability of the indicators generated by the morphological classifier is also degraded by the regularity of the RR intervals. In this case, only the P-wave analyzer classifier and the morphological classifier are relevant for generating the indicators.

[0177] The same applies when a derivation does not allow reliable indicators to be obtained, for example because it is noisy or has artifacts, for example due to muscle movement or electrode detachment, the indicators generated from this derivation, in particular by the P-wave analyzer classifier from combinations of parts of this derivation will not be reliable.

[0178] In order to improve the quality of the global indicators generated by the global classifier, the system advantageously includes a selector INT, SE allowing the selection and / or de-selection of at least one of the indicators generated by the elementary classifiers so that only a selection of these indicators is taken into account by the global classifier GC, to generate the global indicators IGFAk, IGFLAk, IGRSk relating to the time window Fk considered.

[0179] For this purpose, at least one of the indicators, for example each indicator, is advantageously able to be alternately in a selected state and in a deselected state.

[0180] The indicator selection includes each indicator in the selected state. It may include no indicators when all indicators are in the deselected state or include one or more selected indicators.

[0181] The GV generator of the input vector V is configured to generate an input vector V comprising; among the indicators generated by the elementary classifiers for arrhythmias, only each indicator from the indicator selection.

[0182] The value of the other components of the input vector that can take the values ​​of the other indicators, i.e., reserved for the other indicators, is advantageously fixed at the same predetermined value. This value is, for example, equal to / 2 or represents a probability of 1 / 2. For example, in the case of 10 global indicators, if the selection of indicators includes two indicators, the values ​​of the 8 other components of the vector reserved for the global indicators are fixed at the same predetermined value, for example / 2.

[0183] This helps to limit the impact of unselected indicators on the quality of overall indicators.

[0184] The selected state is, for example, the default state of each indicator. The INT, SE selector then advantageously allows at least one of the indicators to be deselected. Advantageously, but not necessarily, the INT, SE selector also allows the indicator to be selected. This allows the indicator to be selected and deselected.

[0185] Alternatively, the deselected state is, for example, the default state of each indicator. The selector then advantageously allows at least one of the indicators to be selected. Advantageously, but not necessarily, the INT, SE selector also allows the indicator to be deselected.

[0186] In a particular embodiment, the INT, SE selector allows for the individual selection and / or deselection of each indicator from a subset of at least one of the indicators generated by the elementary classifiers. It allows, for example, the individual selection of the different indicators generated by the elementary classifiers.

[0187] In another embodiment, the selector allows several indicators to be selected and / or deselected together.

[0188] The INT, SE selector allows, for example, the joint selection and / or deselection of indicators, by selection and / or deselection of at least one elementary classifier and / or by selection and / or deselection of at least one derivation and / or by selection and / or deselection of at least one arrhythmia and / or by selection and / or deselection of at least one group of elementary classifiers.

[0189] For example, each elementary classifier, each derivation and each arrhythmia is capable of being in a selected state and in a deselected state.

[0190] The selected state is, for example, the default state. Alternatively, the deselected state is the default state.

[0191] According to an exemplary embodiment, the selector includes a first selector for individually deselected and / or selected derivation(s) Dk of the time window Fk, such that each indicator generated from the deselected derivation by a predetermined set of elementary classifiers is in the deselected state. This avoids the global classifier GC taking into account the deselected derivation when generating global indicators for the time window Fk.

[0192] The predetermined set of elementary classifiers is, for example, made up of all the elementary classifiers. Thus, deselection of a derivation entails the deselection of each indicator generated from the derivation.

[0193] Alternatively, the predetermined set of elementary classifiers includes a subset of elementary classifiers.

[0194] In the example in [Fig. 3], the system includes at least one elementary classifier, namely MO_FA, MO_FLA, PW_FLA, PW_FA, configured to generate several indicators from their respective derivations, each indicator being generated from a single derivation Dki, Dk2. The predetermined set includes, for example, only these classifiers. This prevents the RR analyzer classifier from being disabled when it uses this derivation.

[0195] Advantageously, as in the example in [Fig. 3], the classifiers of the classifier sets formed for the different arrhythmias are grouped. The classifiers in the same group use the same descriptors from the same derivation(s) to generate their indicators. For example, the morphological classifiers MO_FA and MO_FLA belong to the same group, just as the RR analyzer classifiers belong to the same second group and the P-wave analyzer classifiers belong to the same third group.

[0196] According to an alternative or complementary embodiment, the INT, SE selector includes a second selector allowing for the individual selection and / or deselection of groups of classifiers from the sets of elementary classifiers so that each indicator generated from the deselected group is in the deselected state.

[0197] According to an alternative or complementary embodiment, the INT, SE selector includes a third selector allowing the elementary classifiers of the sets of elementary classifiers dedicated to the respective arrhythmias to be selected and / or de-selected individually so that each indicator generated from each deselected classifier is in the deselected state.

[0198] According to another complementary or alternative example, the INT, SE selector includes a fourth selector allowing the arrhythmias to be selected and / or deselected individually so that each indicator generated for each selected arrhythmia is deselected.

[0199] The different selectors may be identical or different.

[0200] Advantageously, as can be seen in [Fig.1], the processing system S includes an INT user interface.

[0201] The user interface 120 allows a user to enter data or commands so as to be able to interact with the programs according to the invention.

[0202] The INT user interface includes, for example, an INTS output interface and an INTE input interface.

[0203] The input interface includes, for example, a keyboard or a pointing interface, such as a mouse, a light pen, a touchpad, a remote control, a speech recognition device, a haptic device.

[0204] The INTS output interface is designed to provide information to a user, either sensorially or electrically, such as, for example, visually or audibly. The output interface includes, for example, a graphical interface.

[0205] The output interface INTS can be the input device INTE, for example, in the case of a touch tablet.

[0206] Advantageously, the INT user interface allows a user to select and / or deselect at least one of the elementary indicators. In other words, the selector function is performed by the INT user interface.

[0207] The INT user interface allows the user to interact with the classifier C via a CO control module, which is a functional building block of the S system, for the generation of the vector V.

[0208] The INTS output interface is, for example, configured to present to the user a list of indicators that can be selected and / or deselected and / or a list of one or more derivations that can be selected and / or deselected and / or a list of groups of individual classifiers that can be selected and / or deselected and / or at least a list of elementary classifiers that can be selected and / or deselected.

[0209] The INTE input interface allows receiving at least one user selection or deselection command, of an indicator and / or a derivation and / or a group of classifiers and / or an individual classifier taken from the displayed list(s).

[0210] The INT user interface interacts with the vector generator GV, via the CO control module, to generate the input vector V corresponding to the user's command(s). The CO control module is configured to generate, from the user's selection and / or deselection commands, a command to the vector generator GV so that it generates the input vector conforming to the user's commands.

[0211] Advantageously, the INT user interface is configured to allow the selective display, on a screen of the INTS output interface, of at least one indicator and / or at least one global indicator and / or the sequence of states and / or at least one of the derivations and / or at least one of the descriptors used by at least one elementary classifier. This helps the user define which indicators they wish to select and which they wish to deselected.

[0212] Alternatively or in addition, the selector includes an automatic selector SE, which is a functional component of the S system, configured to: • Check if a selection or deselection condition for at least one indicator is met, • automatically select or respectively deselect said at least one indicator when the selection or respectively deselection condition is met.

[0213] For example, the automatic selector SE includes a feature analyzer of the time window derivation(s) configured to check whether the features satisfy a selection or deselection condition of at least one indicator is met and a command generator, which can be implemented by the control module C, configured to generate a selection or deselection command to the vector generator GV so that it generates the input vector V.

[0214] The selection or deselection condition is, for example, a composite condition.

[0215] For example, the characteristic analyzer is configured to check, for each derivation, whether a set of at least one quality criterion is met, including, for example, a first criterion according to which a signal-to-noise ratio is greater than a predetermined threshold (allowing verification of the quality of the derivation), and / or a second criterion according to which the absolute value of the average amplitude of the derivation is greater than a predetermined threshold (allowing verification that the signal is indeed detected).

[0216] The SE automatic selector is configured to deselect each branch that does not meet the quality criterion.

[0217] The automatic selector SE interacts with the vector generator GV, via the control module CO, to generate the input vector V corresponding to the user's command(s). The control module CO is configured to generate, from the selection and / or deselection commands of the automatic selector SE, a command to the vector generator GV so that it generates the input vector conforming to the automatic selector's command and, optionally, to the user's command.

[0218] The first, second, and third selectors can be implemented by the same means, for example, the INT user interface or the automatic selector, or by separate means. For example, one of the selectors is implemented by the INT user interface and the others by the automatic selector.

[0219] Alternatively, the system is devoid of a selector.

[0220] The various functional building blocks mentioned above, which could be artificial neural networks, can more generally be learning functions whose parameters can be fixed by learning during a training phase. Once the training phase is complete, the parameters are fixed and the building blocks are said to be in the trained state.

[0221] These functional building blocks are, for example, in the driven state, in the S system. Process

[0222] The invention also relates to a method for characterizing a heart rhythm. The S system and the SYS system are configured to implement the method according to the invention.

[0223] The steps of the process according to the invention are shown in [Fig.5].

[0224] Advantageously, but not necessarily, the method according to the invention includes a step of acquiring the set of at least one derivation Dk over the time window Fk. This step is implemented by the ACQ acquisition system.

[0225] The method also includes a step 15 of generating descriptors from the set of at least one derivation Dk;.

[0226] This step is advantageously implemented using the GD descriptor generator.

[0227] Alternatively, the generation of the descriptors is carried out prior to the process according to the invention. The process then begins with a step of receiving these descriptors by the system S.

[0228] Step 15 includes a step 151 of detection and time-stamping of beats, that is to say to determine the times of occurrence or TOC, acronym for the Anglo-Saxon expression, of consecutive R waves of at least one of the derivations.

[0229] Step 15 of descriptor generation also includes the generation 160 of statistical indicators characterizing the distribution of the patient's RR intervals during the acquisition window Fk from at least one derivation of the set of at least one derivation Dk;.

[0230] This step is advantageously implemented by running the GIS statistical indicator generator.

[0231] Step 160 advantageously includes generation 161 of an RR series, SRRk for the time window Fk, from at least one derivation of the derivation set, for example derivation Dk;.

[0232] This step includes, for example, the construction of an RR series by calculating the RR intervals between consecutive beats, from the timestamps of consecutive R waves.

[0233] It is advantageously implemented by running the GRR serial generator.

[0234] Stage 160 advantageously includes generation 162, from the RR series, of at least one statistical indicator Gk characterizing the statistical distribution of the intervals R-Rk(h) of the series RR, SRRk.

[0235] This step is advantageously implemented by running the RR GRR serial generator.

[0236] This step advantageously includes the parameterization of a Gaussian mixture model and the calculation of at least one elementary statistical indicator from at least one parameter of the Gaussian mixture model parameter.

[0237] Step 161 advantageously includes the generation of the sequence of ratios between consecutive RR intervals.

[0238] It then advantageously includes, for each arrhythmia, a step 164a, 164b of generation of indicators IFAk, respectively IFLAk, of probabilities of presence of the arrhythmia from the sequence of ratios between RR intervals.

[0239] These steps are advantageously implemented by running the ratio classifiers RR, RA_FA, RA_FLA.

[0240] Step 15 advantageously includes a combination step 152, for example averaging, in which, for each derivation, at least a portion of a subset of the beats present on derivation Dki, respectively Dk2 and detected by the wave detector R is combined, for example averaging.

[0241] During this combination, whole beats or portions of beats can be combined as explained previously.

[0242] This step is implemented by running the MO combiner.

[0243] Step 15 advantageously includes a portion extraction step 153, in which a portion Pki, Pk2 is extracted from one of the combinations or averages Mkb Mk2, obtained from one of the derivations Dkb Dk2. This part includes a portion preceding the R wave of the combination obtained.

[0244] It may correspond to the entire combination obtained or to an incomplete part of the combination obtained.

[0245] For example, this part comprises only a portion of the combination or average preceding Fonde R and devoid of the wave R.

[0246] This portion is advantageously capable of including Fonde P in the case of sinus rhythm.

[0247] This portion, for example, a predetermined duration part preceding Fonde R as explained previously.

[0248] The method then includes a step 20 of generating probability indicators of the presence of arrhythmias from a predetermined set of arrhythmias, during the acquisition window Fk from descriptors of the set of at least one derivation Dk;. This step is implemented using, i.e. by running the classifier C.

[0249] This step 20 includes, in a step 21, the generation, for each arrhythmia in the plurality of arrhythmias, of a set of indicators of the probability of the presence of the arrhythmia over the time window, using a set of classifiers elementary classifiers use descriptors. The elementary classifiers use distinct respective descriptors.

[0250] In a particular embodiment, step 21 comprises the following steps, implemented for each arrhythmia: • generate, during a step 210_FA, respectively 210_FLA, a first set of indicators IMFAki, IMFAk2, respectively IMFLAki, IMFLAk2 of the probability of presence of the arrhythmia AF; respectively FLA, in the time window Fk using a first classifier, which is a morphological classifier MO_FA, respectively MO_FLA, using values ​​from the set of at least one derivation Dkb Dk2 • generate, during a step 211_FA, respectively 211_FLA, a second set of probability indicators for the presence of the arrhythmia IPFAi, IPFA2; respectively IPFLAi, IPFLA2, in the time window Fk, using a second classifier, called the P-wave analyzer classifier PW_FA; respectively PW_FLA, using the first portions including a part preceding the R wave, of combinations of second portions of beats from the set of at least one lead Dkh Dk2, • generate, during a step 212_FA, respectively 212_FLA, a third set of indicators IRRFAk; respectively IRRFLAk, of probabilities of presence of the arrhythmia in the time window Fk, using a third classifier called RR interval analyzer classifier, RR_FA; respectively RR_FLA, from the set of statistical indicators representative of the R_R interval distribution of the electrocardiogram ECGk.

[0251] Step 212_FA, respectively 212_FLA is advantageously implemented by running the interval analyzer classifier RR, RR_FA; respectively RR_FLA from a vector composed of at least one statistical indicator generated by the GIS generator, for example the indicator Gk characterizing the R_R distribution and the probability indicator generated for the arrhythmia considered IFAk, IFLAk.

[0252] In the case of multiple derivations, at least one classifier uses both derivations as described above. During its execution, it then generates different indicators for the two derivations. Advantageously, it processes the derivations consecutively, so as to successively generate the IMFAk1 and IMFAk2 indicators. This is, for example, the case for the first and second classifiers.

[0253] The method then advantageously includes a step 40 of generating an input vector Vk, by running the input vector generator using itself the sets of probability indicators generated during step 21 for the time window Fk.

[0254] The method then advantageously includes a step 50 of generating global indicators IGFAk, IGFLAk, of the probability of the presence of arrhythmias in the time window Fk from the input vector Vk. This step is implemented by running the global classifier GC.

[0255] Advantageously, the global indicators include an IGRSk indicator of the probability of the presence of sinus rhythm in the Fk time window.

[0256] Advantageously, the preceding steps are repeated so as to generate sequences of K sets of global indicators IGFAk, IGFLAk, IGRSk from set descriptors of at least one successive derivation Dkb Dk2.

[0257] In other words, the process includes the acquisition of a sequence of sets of at least one derivation, i.e. of ECGk with k= 1 to K, acquired during K successive time windows F, for example consecutive, of a global acquisition window FG of global duration TG, the generation of descriptors from the set of at least one derivation acquired for each of these windows.

[0258] Alternatively, the method only includes receiving sets of descriptors respectively associated with sets of at least one successively acquired derivation, that is, acquired over successive time windows Fk, and implementing steps 21, 40, and 50 for each of these sets of descriptors. Each set of global indicators IGFAk, IGFLAk, IGRSk is generated from a single set of at least one derivation Dki, Dk2 of the sequence of time windows Fk. The temporal order of the sets of global indicators IGFAk, IGFLAk, IGRSk is the same as that of the derivation sets DkB and Dk2 from which they are respectively generated.

[0259] Advantageously, the method includes a time analysis step 60, to define a sequence of Ek states of the heart rhythm corresponding to the Ek states of the heart rhythm on successive time windows Fk from the sequence of global indicator sets IGFAk, IGFLAk, IGRSk generated for successive time windows Fk.

[0260] Advantageously, this step is implemented by running the TC time analyzer from the global indicator sets IGFAk, IGFLAk, IGRSk generated for successive time windows.

[0261] Advantageously, the method includes, for at least one time window Fk, a step 30 of selecting and / or deselected at least one of the indicators generated during step 21 for the corresponding time window Fk, such that the indicator selection is composed of each selected indicator. The vector Vk is generated, during step 40, from the indicator selection.

[0262] This step 30 is implemented by the selector.

[0263] It is, for example, implemented by means of the human-machine interface and / or by means of the automatic selector.

[0264] In the case of the presence of the selection and / or deselection step 30, for a time window, the method may include step 40 of generating the input vector using the vector generator with each indicator in the selected state prior to the selection / deselection, as well as step 50 of generating the global indicators from the input vector and step 60 of temporal analysis from the global indicators generated for the time windows, including window Fk. The method advantageously includes updating the input vector, the global indicators, and the set of states by repeating steps 40, 50, and 60 based on the state selection obtained following the selection step 30.

[0265] The training of the drive functions used in the process according to the invention is advantageously implemented prior to the process according to the invention. It is advantageously supervised, at least for the classifiers.

[0266] Advantageously, the input data of at least one classifier, for example, of each of the classifiers, are calibrated prior to their use by the classifier so that they fall within predetermined value ranges. The system advantageously includes calibration means configured for these functions. And the method advantageously, but not necessarily, includes these steps.

[0267] By way of example, the SYS system includes means for calibrating the derivations so that the derivation values ​​used by the morphological classifier are calibrated values ​​of these derivations, i.e., values ​​within a predetermined range. The analog converter can, for example, perform this function.

[0268] Advantageously, the training data are calibrated in the same way. Hardware

[0269] From a hardware point of view, the processing system S can be seen as a calculator interacting with a computer program product.

[0270] The system S includes at least one computer, for example, a microcomputer, a network of computers, an electronic component, a tablet, a Smartphone or a personal digital assistant (PDA).

[0271] The MEM memory assembly includes, for example, a computer-readable medium. The computer-readable medium is a tangible device readable by the reader of the computing unit, capable of storing electronic instructions and of being coupled to communication means of the S system.

[0272] In other words, the computer-readable medium is a tangible medium. In other words, it is not a transient signal in itself, such as radio waves or other waves Freely propagating electromagnetic waves, such as light pulses or electronic signals. Such a computer-readable storage medium is, for example, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof.

[0273] By way of example, the readable medium is an optical disc, a magneto-optical disc, a read-only memory (ROM), an erasable and programmable read-only memory (EPROM), an electrically erasable and programmable read-only memory (EEPROM), a random access memory (RAM), a magnetic card, or an optical card. The memory assembly may include an operating system and load the programs according to the invention. It includes registers adapted to store parameter variables created and modified during the execution of the aforementioned programs. A computer program containing software instructions is then stored on the readable medium.

[0274] The data processing unit TR includes at least one electronic circuit designed to manipulate and / or transform data represented by electronic or physical quantities in registers of the evaluation system and / or memories into other similar data corresponding to physical data in register memories or other types of display devices, transmission devices or storage devices.

[0275] The data processing unit TR includes, for example, memories for storing data, operationally coupled to the data processing circuit and a reader adapted to read the computer-readable media.

[0276] Each function of the system S is executed by causing the data processing unit TR to read a predetermined program on hardware such as the MEM memory set so that the data processing circuit performs calculations, controls communications and reads and / or writes data into the MEM memory set.

[0277] The process is carried out on a single computer or on a system distributed among several computers (in particular via the use of cloud computing).

[0278] The data processing unit comprises at least one of the following: a set of one or more processors (for example, a central processing unit (CPU), a graphics processing unit (GPU), a microcontroller and a digital signal processor (DSP)) capable of interpreting instructions in the form of a computer program; a programmable logic circuit, such as an application-specific integrated circuit (ASIC), or a network of programmable gates. in situ (FPGA), a programmable logic device (PLD) of programmable logic arrays (PLA), a system on chip (SOC), and / or an electronic board in which steps of the process according to the invention are implemented in hardware elements.

[0279] The product-program may include computer-readable recording media.

[0280] Alternatively, the program instructions are obtained from an external source and downloaded via a network. This is particularly the case for applications. In this case, the computer program product includes a computer-readable data carrier on which the program instructions are stored or a data carrier signal on which the program instructions are encoded.

[0281] The invention relates to a computer program product comprising the computer-readable medium containing instructions which, when executed by the processing circuit, cause the system S to implement the steps of the process according to the invention, that is to say, to execute the functional building blocks of the system according to the invention.

[0282] The form of program instructions is, for example, a form of source code, a computer-executable form, or any intermediate form between source code and a computer-executable form, such as the form resulting from the conversion of source code via an interpreter, assembler, compiler, linker, or locator. Alternatively, program instructions are microcode, firmware instructions, state definition data, integrated circuit configuration data (e.g., VHDL), or object code. Program instructions are written in any combination of one or more programming languages, for example, an object-oriented programming language (C++, Java, Python), a procedural programming language (e.g., C).

[0283] The means of communication enable communication between the elements of the system S and possibly between at least one element of the system and a device external to the system S. The means of communication can establish a physical link between elements of the system and / or between an element of the system and a device external to the system and / or a remote (wireless) communication link between elements of the system and / or between an element of the system and a device external to the system.

[0284] The means of communication advantageously facilitate data transmission between the ACQ acquisition system and the S system. It allows, for example, communication between the box and the classifier C in the associated embodiment.

Claims

Demands

1. A cardiac rhythm characterization system comprising a data processing unit configured to implement, by computer, the following steps: • For each arrhythmia in a plurality of arrhythmias, generate, using a set of elementary classifiers, a set of probability indicators for the presence of the arrhythmia over a time window using descriptors from at least one derivation of a patient's electrocardiogram acquired over the time window, at least one indicator being capable of being alternately in a selected and deselected state, • generate an input vector (Vk) from the indicator sets generated for the arrhythmias, the components of the input vector (Vk) comprising, from among the indicators in the indicator sets generated for the arrhythmias, only each indicator from a selection of indicators, • generate,using a global classifier (GC), global indicators of the probability of presence of arrhythmias in the time window (Fk) from the input vector (Vk), the system including a selector (INT, SE) allowing selection and / or deselection of at least one of the indicators so that the indicator selection is composed of each selected indicator.

2. System according to the preceding claim, wherein the value of each component of the input vector (Vk) associated with one of the global indicators is fixed to the same predetermined value.

3. A system according to any one of the preceding claims, wherein the selector allows for the individual selection and / or de-selection of each indicator from a subset of at least one of the indicators generated using the elementary classifiers of the sets of elementary classifiers.

4. A system according to any one of the preceding claims, wherein the set of at least one derivation comprises several derivations and wherein the selector comprises a first selector enabling the derivations to be deselected and / or selected individually, such that each first indicator of the indicator sets generated, using each respective elementary classifier of a predetermined subset of the elementary classifiers, from a deselected derivation is deselected.

5. System according to the preceding claim, wherein the subset consists of each elementary classifier configured to generate first indicators of the probability of presence of arrhythmias from first descriptors of the set of at least one derivation, each first indicator being generated from a single derivation.

6. A system according to any one of the preceding claims, wherein the selector includes a second selector for individually selecting and / or deselecting groups of elementary classifiers from sets of elementary classifiers such that each indicator generated using a selected group is in the deselected state, the groups of elementary classifiers using respective descriptors distinct from the set by at least one derivation, the elementary classifiers of the same group using the same descriptors.

7. A system according to any one of the preceding claims, wherein the selector includes a third selector for individually selecting and / or deselecting elementary classifiers from sets of classifiers such that each indicator generated using a selected elementary classifier is in the deselected state.

8. A system according to any one of the preceding claims, comprising a user interface (INT) enabling a user to select and / or deselect at least one of the elementary indicators.

9. A system according to any one of the preceding claims, wherein the selector comprises an automatic selector (ES) configured to: • Check whether a selection or deselect condition of at least one indicator is met, • automatically select or respectively deselect said at least one indicator when the selection or respectively deselect condition is met.

10. A system according to any one of the preceding claims, wherein the generation of the set of arrhythmia probability indicators comprises: • generating a first set of arrhythmia probability indicators (IMFAki, IMFAk2) in the time window using a first classifier using values ​​from the set of at least one lead (Dki, Dk2), • generating a second set of arrhythmia probability indicators in the time window, using a second classifier from first portions, preceding the R wave, of combinations of second portions of beats from the set of at least one lead,• generate a third set of probability indicators for the presence of arrhythmia within the time window using a third classifier with a set of statistical indicators representative of the R_R interval distribution of the electrocardiogram.

11. A system according to any one of the preceding claims, wherein the data processing unit is configured to define a sequence of states (Ek) of the individual's heart rhythm, taken from arrhythmias and sinus rhythm, from sets of global indicators generated for successive time windows.

12. A system according to the preceding claim, wherein the definition of the state sequence uses a Viterbi algorithm configured to determine the most probable sequence of states likely to be obtained by a hidden Markov model from global indicator sets.

13. System according to any one of the preceding claims comprising an acquisition device (ACQ) comprising electrodes configured to acquire the entirety of at least one lead of the electrocardiogram and a processing subunit configured to use a descriptor generator to generate the descriptors of the entirety of at least one lead.

14. System (SYS) according to any one of the preceding claims, wherein the arrhythmias are atrial fibrillation and atrial flutter.

15. A system according to any one of the preceding claims, wherein each of the elementary classifiers and the global classifier are trained classifiers.

16. A method for characterizing a cardiac rhythm comprising: • for each arrhythmia in a plurality of arrhythmias, generating, using a set of elementary classifiers, a set of probability indicators for the presence of the arrhythmia over a time window using descriptors from at least one lead of a patient's electrocardiogram acquired over the time window, the indicators being capable of being alternately in the selected and deselected states, • generating an input vector (Vk) from the sets of indicators generated for the arrhythmias, the components of the input vector (Vk) comprising, from among the indicators in the sets of indicators generated for the arrhythmias, only each indicator from a selection of indicators, • generating, using a global classifier,global indicators of the probability of arrhythmias occurring within the time window, based on the input vector (Vk), select or deselect at least one of the indicators so that the indicator selection is composed of each selected indicator.

17. Product computer program comprising a readable information carrier, on which is stored a computer program comprising program instructions, the computer program being loadable onto a data processing unit and adapted to drive the implementation of steps of the process according to the preceding claim, when the computer program is implemented on the data processing unit.

18. Readable information carrier comprising program instructions forming a computer program, the computer program being loadable onto a data processing unit and adapted to drive the implementation of a step of the process according to claim 16 when the computer program is implemented on the data processing unit.

Images