An ECG Morphological Analysis Algorithm for Hybrid Patient Monitoring

Abstract

Screening and diagnosis in ubiquitous healthcare environments is performed by the medical staff in the presence of a variety of medical data originating from continuous in vivo patient monitoring. This assumes the correlation of data, such as vital signs, biopotentials, activity tracking, bio-markers, etc., in the context of hybrid clinical monitoring. The hybrid patient monitoring setup envisioned in this work targets the simultaneous assessment of body fluids via electro-optical means and of electrophysiological data, i.e. electrocardiogram, via biopotential monitoring. Accordingly, the aim of this paper is to propose an electrocardiogram morphological analysis algorithm applicable in the framework of the envisioned hybrid monitoring setup. Electrocardiogram morphological analysis is performed in time domain in two stages. The R, S and Q waves are first identified via comparison to a threshold value. Then, the P and T waves are identified via cross-correlation to a reference signal. The novelty of this work is that cross-correlation is performed after the QRS complex was eliminated from the electrocardiogram. Expectedly, this leads to improved results in wave identification and to a reduction of false detections. Matlab simulation results validate the proposed procedure.