A Fast Machine Learning Model for ECG-Based Heartbeat Classification and Arrhythmia Detection

no code implementations • Frontiers in Physics 2019 • Miquel Alfaras, Miguel C. Soriano, Silvia Ortín

In the MIT-BIH AR database, our classification approach provides a sensitivity of 92. 7% and positive predictive value of 86. 1% for the ventricular ectopic beats, using the single lead II, and a sensitivity of 95. 7% and positive predictive value of 75. 1% when using the lead V1'.

Arrhythmia Detection BIG-bench Machine Learning +3