Apparatus for treating cardiac arrhythmias utilizing a machine learning algorithm to optimize an ablation index calculation
A method is provided. The method includes receiving, by an optimization engine, inputs from previous ablation procedures. The method also includes training, by the optimization engine, a machine learning algorithm of the optimization engine to learn scenarios for the previous ablation procedures. The method also includes generating, by the optimization engine, ablation indices for the scenarios.
1 . A method for performing a cardiac ablation that improves treatment of cardiac arrhythmias, the method comprising:
receiving a plurality of inputs from one or more previous ablation procedures;
training a machine learning algorithm based on the plurality of inputs to generate a trained model, wherein the plurality of inputs comprises data representing an amount of contact force and energy applied to tissue in the or more previous ablation procedures;
receiving, during treatment of a patient, real-time data comprising synchronised body-surface electrocardiogram (BS-ECG) signals and intracardiac electrogram (IC-ECG) signals detected from cardiac tissue of the patient;
determining at least one ablation index value by
encoding the BS-ECG and IC-ECG signals with a deep-learning encoder that forms part of the trained model to produce a latent representation of beat-to-beat conduction properties, and
combining the latent representation with data representing an amount of contact force and energy to compute the cardiac at least one ablation index value;
configuring a catheter to perform the cardiac ablation based on the at least one ablation index value; and
controlling the catheter to administer the cardiac ablation.
2 . The method of claim 1 , wherein the plurality of inputs further comprises electromagnetic properties applied during the previous ablation procedures, and a time for the previous ablation procedures.
3 . The method of claim 1 , wherein the plurality of inputs further comprises ablation continuous normalized input parameters and ablation categorical normalized input parameters.
4 . The method of claim 3 , wherein the ablation continuous normalized input parameters comprises ablation time duration, initial impedance, radial location, time duration between current ablation to former ablation, catheter stability, ablation power, ablation maximum temperature, distance from former ablation, or impedance reduction parameters.
5 . The method of claim 3 , wherein the ablation categorical normalized input parameters comprise left atria right side, left atria left side, pulmonary veins ostia, tissue character, age class, or background disease.
6 . The method of claim 1 , wherein the machine learning algorithm is unsupervised or supervised.
7 . The method of claim 1 , wherein the plurality of inputs are stored by a database that is remote from the patient during the treatment of the cardiac arrhythmias.
8 . A system for performing a cardiac ablation that improves treatment of cardiac arrhythmias, the system comprising:
a memory;
a catheter; and
one or more processors communicatively coupled to the memory and the catheter, wherein the one or more processors are collectively configured to:
receive, from the memory, a plurality of inputs from one or more previous ablation procedures;
train a machine learning algorithm to generate a trained model based on the plurality of inputs, wherein the plurality of inputs comprises data representing an amount of contact force and energy applied to tissue in the previous ablation procedures;
receive, during treatment of a patient, real time data comprising synchronised BS-ECG signals and IC-ECG signals detected from cardiac tissue of the patient;
determine at least one ablation index value by
encoding the BS-ECG and IC-ECG signals with a deep-learning encoder that forms part of the trained model to produce a latent representation of beat-to-beat conduction properties, and
combining the latent representation with data representing an amount of contact force and energy to compute the at least one ablation index value;
configure the catheter to perform the cardiac ablation based on the at least one ablation index value; and
control the catheter to administer the cardiac ablation.
9 . The system of claim 8 , wherein the plurality of inputs further comprises electromagnetic properties applied during the previous ablation procedures, and a time for the previous ablation procedures.
10 . The system of claim 8 , wherein the plurality of inputs further comprises ablation continuous normalized input parameters and ablation categorical normalized input parameters.
11 . The system of claim 10 , wherein the ablation continuous normalized input parameters comprises ablation time duration, initial impedance, radial location, time duration between current ablation to former ablation, catheter stability, ablation power, ablation maximum temperature, distance from former ablation, or impedance reduction parameters.
12 . The system of claim 10 , wherein the ablation categorical normalized input parameters comprise left atria right side, left atria left side, pulmonary veins ostia, tissue character, age class, or background disease.
13 . The system of claim 8 , wherein the machine learning algorithm is unsupervised or supervised.
14 . The system of claim 8 , wherein the memory is a database that is remote from the patient during the treatment of the cardiac arrhythmias.
15 . A non-transitory computer readable storage medium storing instructions for performing a cardiac ablation that improves treatment of cardiac arrhythmias, the instructions when executed by one or more processors of a console cause the console to perform a method comprising:
receiving a plurality of inputs from one or more previous ablation procedures;
training a machine learning algorithm based on the plurality of inputs to generate a trained model, wherein the plurality of inputs comprises data representing an amount of contact force and energy applied to tissue in the previous ablation procedures;
receiving, during treatment of a patient, real-time data comprising synchronised body-surface electrocardiogram (BS-ECG) signals and intracardiac electrogram (IC-ECG) signals detected from cardiac tissue of the patient;
determining at least one ablation index value by
encoding the BS-ECG and IC-ECG signals with a deep-learning encoder that forms part of the trained model to produce a latent representation of beat-to-beat conduction properties, and
combining the latent representation with data representing an amount of contact force and energy to compute the at least one ablation index value;
configuring a catheter to perform the cardiac ablation based on the at least one ablation index value; and
controlling the catheter to administer the cardiac ablation.