ECG Signal Dataset

This dataset contains electrocardiogram (ECG) signal recordings and extracted physiological parameters, collected for the purpose of classifying heart conditions into two categories: Normal and Arrhythmia.

Data was collected from 30 male university students, aged 18 to 22 years, at Universitas Prima Indonesia. Each participant was recorded under three physical activity conditions—sitting, walking, and running—with 3 minutes of ECG data per activity, totaling 9 minutes per subject. The ECG signals were captured using a portable ECG monitoring device, with electrodes placed on the chest to ensure accurate and consistent signal quality. Subjects were instructed to remain relaxed while sitting and to walk or run under controlled supervision.

The raw ECG signals were preprocessed to improve signal quality, including noise removal, normalization, and feature extraction. The resulting dataset includes 11 features for each ECG sample:

RR interval

PR interval

QS interval

QT interval

ST interval

Standard deviations of each interval

Heart rate (in BPM)

These features capture essential aspects of cardiac electrical activity and variability, making the dataset highly relevant for arrhythmia detection, ECG classification, and biomedical signal processing.