Adaptive QRS Detection Based on Maximum A Posteriori Estimation

A mathematical model for the occurrence of nonoverlapping pulse-shaped waveforms corrupted with colored Gaussian noise is considered for the purpose of QRS detection. The number of waveforms, the arrival times, amplitudes, and widths are regarded as random variables. The joint MAP estimation of all the unknown quantities consists of linear filtering followed by an optimization procedure. A class of filters is introduced which is easy to implement. The mismatching obtained by using this class for detection of model QRS complexes is investigated. The optimization procedure is time-consuming and is modified so that a threshold test is obtained. The model formulation with nonoverlapping waveforms leads to an "eye-closing" procedure covering a segment before as well as after an accepted event. Adaptivity of the detector is gained by utilizing past as well as future signal properties in determining thresholds for QRS acceptance.