Reconstruction of a surface electrocardiogram from an endocardial electrogram using non-linear filtering

Patent metadata

Field	Value
Publication number	US-9681815-B2
Application number	US-201314015770-A
Country	US
Kind code	B2
Filing date	Aug 30, 2013
Priority date	Apr 6, 2009
Publication date	Jun 20, 2017
Grant date	Jun 20, 2017

Abstract

The present invention relates to an active medical device that uses non-linear filtering for reconstructing a surface electrocardiogram from an endocardial electrogram. At least one endocardial EGM electrogram signal is collected from of samples collected from at least one endocardial or epicardial derivation ( 71′, 72′, 73 ′), and at least one of a reconstructed surface electrocardiogram (ECG) signal through the processing of collected EGM samples by a transfer function (TF) of a neural network ( 60 ′). The neural network ( 60 ′) is a time-delay-type network that simultaneously processes said at least one endocardial EGM electrogram signal, formed by a first sequence of collected samples, and at least one delayed version of this EGM signal, formed by a second sequence of collected samples distinct from the first sequence collected samples. The neural network ( 60 ′) provides said reconstructed ECG signal from the EGM signal and its delayed version.

First claim

The invention claimed is:

1. A method for reconstructing electrocardiogram signals, the method comprising:
obtaining at least one endocardial electrogram (EGM) signal, formed from a plurality of signal samples from at least one endocardial or epicardial derivation;
obtaining a transfer function; and
generating, with a time-delay type neural network, a first surface electrocardiogram (ECG) signal from the at least one EGM signal by simultaneously processing (A) the at least one EGM signal and (B) at least one delayed version of the of the at least one EGM signal, wherein the time-delay neural network is selected from a group comprising a focused time-delay neural network, a distributed time-delay neural network, and a recurrent time-delay neural network.

2. The method of claim 1 , further comprising adapting the transfer function responsive to the at least one EGM signal.

3. The method of claim 1 , wherein obtaining at least one EGM signal further comprises transferring the at least one EGM signal from an implantable cardiac device.

4. The method of claim 1 , wherein obtaining the transfer function further comprises training the transfer function responsive to a first training signal comprising at least two continuous heartbeats.

5. The method of claim 1 , wherein obtaining the transfer function further comprises training the transfer function responsive the at least one EGM signal and at least one simultaneously collected ECG signal.

6. The method of claim 1 , wherein obtaining the transfer function further comprises training the transfer function responsive to an EGM signal and a plurality of differently delayed versions of the EGM signal.

7. The method of claim 1 , further comprising generating, with a second time-delay type neural network and second transfer function, a second surface ECG signal from the at least one EGM signal by simultaneously processing (A) the at least one EGM signal and (B) at least one delayed version of the of the at least one EGM signal.

8. A system for reconstructing electrocardiogram signals, the system comprising:
an implantable device configured to acquire at least one endocardial electrogram (EGM) signal, formed from a plurality of signal samples from at least one endocardial or epicardial derivation; and
a time-delay neural network having a plurality of neurons and a transfer function, wherein the time-delay neural network generates a surface electrocardiogram (ECG) signal by simultaneously processing (A) the at least one EGM signal and (B) at least one delayed version of the at least one EGM signal, the time-delay neural network is selected from a group comprising a focused time-delay neural network, a distributed time-delay neural network, and a recurrent time-delay neural network.

9. The system of claim 8 , wherein the time-delay neural network is external to the implantable device.

10. The system of claim 8 , wherein the implantable device is configured to store the at least one EGM signal for a predetermined amount of time.

11. The system of claim 8 , wherein the transfer function is trained responsive to a first training signal comprising at least two continuous heartbeats.

12. The system of claim 8 , wherein the transfer function is trained responsive the at least one EGM signal and at least one simultaneously collected ECG signal.

13. The system of claim 8 , wherein the transfer function is trained responsive to an EGM signal and a plurality of differently delayed versions of the EGM signal.

14. The system of claim 8 , further comprising a second time-delay type neural network and second transfer function configured to generate a second surface ECG signal by simultaneously processing (A) the at least one EGM signal and (B) at least one delayed version of the of the at least one EGM signal.

15. A system for reconstructing electrocardiogram signals, the system comprising:
at least one processor;
a computer readable storage device storing instructions therein, the instructions, when executed by the at least one processor, cause the at least one processor to:
obtain at least one endocardial electrogram (EGM) signal, formed from a plurality of signal samples from at least one endocardial or epicardial derivation;
obtain a transfer function; and
generate, with a time-delay type neural network, a first surface electrocardiogram (ECG) signal from the at least one EGM signal by simultaneously processing (A) the at least one EGM signal and (B) at least one delayed version of the of the at least one EGM signal, wherein the time-delay neural network is selected from a group comprising a focused time-delay neural network, a distributed time-delay neural network, and a recurrent time-delay neural network.

16. The system of claim 15 , wherein execution of the instructions further cause the at least one processor to adapt the transfer function responsive to the at least one EGM signal.

17. The system of claim 15 , wherein execution of the instructions further cause the at least one processor to train the transfer function responsive to a first training signal comprising at least two continuous heartbeats.

18. The system of claim 15 , wherein execution of the instructions further cause the at least one processor to train the transfer function responsive the at least one EGM signal and at least one simultaneously collected ECG signal.

19. The system of claim 15 , wherein execution of the instructions further cause the at least one processor to train the transfer function responsive to an EGM signal and a plurality of differently delayed versions of the EGM signal.

20. The system of claim 15 , wherein execution of the instructions further cause the at least one processor to generate, with a second time-delay type neural network and second transfer function, a second surface ECG signal by simultaneously processing (A) the at least one EGM signal and (B) at least one delayed version of the of the at least one EGM signal.

Assignees

• Sorin Crm Sas
• Sorin Crm Sas

Inventors

• Porée Fabienne
• Carrault Guy
• Kachenoura Amar
• Hernandez Alfredo

Classifications

• G16H50/70

  for mining of medical data, e.g. analysing previous cases of other patients · CPC title

• A61B5/7267 Primary

  involving training the classification device · CPC title

• G06F2218/02

  Preprocessing · CPC title

• A61B5/319

  Circuits for simulating ECG signals · CPC title

• G06K9/6232

  Physics · mapped topic