US 12,257,060 B2
Methods and systems for predicting arrhythmia risk utilizing machine learning models
Kevin Davis, Thousand Oaks, CA (US); and Aditya Goil, Stevenson Ranch, CA (US)
Assigned to Pacesetter, Inc., Sylmar, CA (US)
Filed by Pacesetter, Inc., Sylmar, CA (US)
Filed on Dec. 17, 2021, as Appl. No. 17/554,745.
Claims priority of provisional application 63/167,213, filed on Mar. 29, 2021.
Prior Publication US 2022/0304612 A1, Sep. 29, 2022
Int. Cl. A61B 5/00 (2006.01); A61B 5/283 (2021.01); A61B 5/352 (2021.01); A61B 5/353 (2021.01); A61B 5/358 (2021.01); A61B 5/363 (2021.01); A61B 5/366 (2021.01); A61B 5/389 (2021.01)
CPC A61B 5/363 (2021.01) [A61B 5/0006 (2013.01); A61B 5/283 (2021.01); A61B 5/352 (2021.01); A61B 5/353 (2021.01); A61B 5/358 (2021.01); A61B 5/366 (2021.01); A61B 5/389 (2021.01); A61B 5/686 (2013.01); A61B 5/7267 (2013.01); A61B 5/7275 (2013.01); A61B 5/746 (2013.01)] 22 Claims
OG exemplary drawing
 
1. A system for determining an arrhythmia risk, comprising:
memory to store specific executable instructions and a machine learning (ML) model trained to predict an arrhythmia with a characteristic of interest (COI) that exhibits a non-physiologic behavior, the ML model trained based on pseudo cardiac activity (CA) signals, the pseudo CA signals generated by a pseudo signal generator (SG) model based on actual first CA signals;
one or more processors configured to execute the specific executable instructions to:
obtain second CA signals collected by an implantable medical device (IMD), wherein the COI, in the second CA signals collected, exhibits a physiologic behavior; and
apply the ML model to the second CA signals to identify a risk factor that a patient will experience the arrhythmia at a future point in time even though the COI from the second CA signals, exhibits the physiologic behavior.