| CPC G06N 3/08 (2013.01) [A61B 5/287 (2021.01); A61B 5/339 (2021.01); A61B 5/349 (2021.01); A61B 5/7267 (2013.01); G06N 3/04 (2013.01)] | 16 Claims |
|
1. A method, comprising:
collecting a plurality of bipolar electrograms and respective unipolar electrograms of patients, the electrograms comprising annotations in which one or more human reviewers have identified and marked a window-of-interest and one or more activation times inside the window-of-interest;
generating, from the electrograms, a ground truth data set for training at least one electrogram-preprocessing step of a Machine Learning (ML) algorithm, wherein the electrogram-preprocessing step comprises specifying coefficients of a set of convolutional kernels and performing one or more convolutions of the electrograms with the set of convolutional kernels and point-by-point multiplication between a bipolar electrogram filtered by one of the convolutional kernels and a respective unipolar electrogram filtered by another one of the convolutional kernels, wherein applying the ML algorithm comprises inputting a multiplication signal resulting from the point-by-point multiplication to the ML algorithm; and
applying the ML algorithm to the electrograms, to at least train the at least one electrogram-preprocessing step, so as to detect an occurrence of an activation in a given bipolar electrogram within the window-of-interest.
|