The QRS detection is key component of each automated ECG analysis. For this purpose a lot of QRS algorithms have been already developed. In the same time the number of new published methods continues to grow up. This implicitly proves the impossibility of building such detector that could totally cover the variety of all shapes of ventricular beats encountered in practice. Generally, limited studies on discrimination between normal (sinus) and ectopic beats are available. The paper describes very fast procedure for accurate QRS detection in long term ECG Holter recordings, followed by classification of the complexes in normal and ectopic. The algorithm was tested with the widely accepted AHA and MIT-BIH databases. The obtained sensitivity and specificity are comparable to other published results.

, , ,
International Journal Bioautomation
Sprott School of Business

Tanev, S. (2012). Ventricular beat detection and classification in long term ECG recordings. International Journal Bioautomation, 16(4), 273–290.