QRS complex detection based on primitive

The detection of the QRS complex is one of the most important issues in electrocardiogram (ECG) signal analysis. Although research on the detection of the R-peak has demonstrated a high detection rate through a diverse number of studies, research on the detection of the onset and offset boundaries of the QRS complex has proven to be difficult, as the locations of these endpoints are often unclear, and the detection results are difficult to interpret. Hence, detection research through improved algorithms continues to be an important component of the ECG signal analysis, especially given the importance of the QRS complexs role in the diagnosis of arrhythmia through measuring the length of the onset and offset of the QRS complex. This paper proposes an improved algorithm that focuses on the primitive of the QRS complex for detecting the onset and offset of the complex based on the morphological characteristics of the QRS complex. The proposed algorithm was tested through experiments based on QT database (QT-DB) data provided by Physionet, and the outcome revealed not only the reliable detection of the QRS complex boundaries but also results that were superior to the location information recorded in the QT-DB.