Discrimination between RA and LA Sinus Rhythms using machine learning approach.

Area of Science:

• Cardiology
• Biomedical Engineering
• Machine Learning

Background:

• Atrial fibrillation (AF) is a prevalent cardiac arrhythmia with potentially fatal outcomes.
• Machine learning (ML) is utilized for classifying ECG signals, distinguishing AF from sinus rhythm post-ablation.
• Intracardiac electrograms (iEGMs) from the left atrium (LA) and right atrium (RA) may exhibit distinct sinus rhythm characteristics.

Purpose of the Study:

• To develop a method for evaluating iEGMs in a high-dimensional feature space.
• To effectively discriminate between sinus rhythms recorded from the LA and RA.
• To investigate the similarity in feature space distribution between LA rhythms post-ablation and sinus RA rhythms.

Main Methods:

• Feature extraction from time-series iEGMs.
• Application of Support Vector Machine (SVM) and K-means clustering algorithms.
• Evaluation of classification accuracy for distinguishing LA and RA iEGMs.

Main Results:

• A method was demonstrated to effectively discriminate between LA and RA sinus rhythms using iEGMs.
• Non-supervised K-means clustering achieved 90.15% classification accuracy.
• Rhythms in the LA post-ablation showed a similar distribution in feature space to sinus RA rhythms.

Conclusions:

• The study successfully differentiated baseline iEGMs from the LA and RA.
• The findings provide valuable insights into signal identification using iEGMs.
• The developed method shows promise for improved arrhythmia analysis.