Algorithm for automatic beat detection of cardiovascular pressure signals

Pressure beat detection is an integral part of most analysis techniques for arterial blood pressure (ABP), intracranial pressure (ICP), and pulse oximetry (SpO2) signals. Beat detection has been used to estimate heart rate in the ABP signal, to classify ICP morphologies, and to estimate blood pressure using pulse oximeter waveforms. This paper describes an algorithm that was developed to detect pressure peak beats in ABP, ICP, and SpO2 signals. When compared to the expert annotation of several signals consisting of over 42,500 pressure beats, the algorithm detected pressure peaks with an average sensitivity of 99.6% ± 0.27 and an average positive predictivity of 98.6% ± 1.1.