MIT-BIH arrhythmia dataset

The MIT-BIH Arrhythmia Database, co-published by the Arrhythmia Laboratory at Beth Israel Hospital in Boston and MIT, contains 48 half-hour dual-channel ambulatory electrocardiogram (ECG) recordings taken at a sampling rate of 360 Hz from 47 subjects, totaling more than 110,000 individual heartbeats independently labeled by two or more cardiologists, and the dataset considers only the the first ECG lead, Modified Limbic Lead II (MLII). The initial 15 ECGs were categorized into five categories, N, S (SVEB), V (VEB), F, and Q, according to the recommendations of the American Association for the Advancement of Medical Instrumentation (AAMI), a standard that primarily recommends performance evaluation for the between-patient scenario. Therefore, we categorized the 44 signal recordings (excluding 102, 104, 107, and 217) into DS1 and DS2 according to the between-patient paradigm for training and testing, respectively