Table 1.

Machine learning methods and corresponding applications in the detection and management of atrial fibrillation

Machine learning method	Description	Example application	Reference
Traditional machine learning
Cox regression	Probability distribution estimating time to a pre-specified event	Prediction of post-ablation AF recurrence	[55]
Support vector machine	Utilizes hyperplane to separate two classes non-linearly	AF detection through HRV analysis of photoplethysmography readings	[23]
Random forest	Average of hierarchical decision trees’ interpretation	Locating re-entrant drivers in AF	[56]
Deep learning
Convolutional neural network	Mimics biological neural networks by incorporating nodes processing data in a hierarchical fashion	Detection of AF from a sinus-rhythm 12-lead ECG	[32]

AF atrial fibrillation, HRV heart rate variability