Assessment of pacemaker function in ambulatory ECG recordings using real time analysis

We describe a digital PM software module for automatic analysis of Pace-Maker (PM) function, developed from an experimental real-time ambulatory ECG analyzer. PM spikes are defected by means of a time/slope-sensitive algorithm, working on an high-pass filtered ECG signal, sampled at 1000 sps. The occurrence of each spike is recorded, the spike artefact is masked and then the ECG signal is digitally low-pass filtered and rescanned for QRS detection and classification. Both the compressed ECG signal and the real-time analysis results are stored on a solid state memory for off-line review. The number of spike(s) and the time relations between QRS and spikes are evaluated to classify them. The pacing mode as well as many of the PM parameters are automatically assessed by analyzing the distributions of the spike-spike, R to spike and spike to R intervals. The accuracy of our analyzer was assessed on an annotated data-base of 20 PM rhythms and confirmed by the analysis of 45 other unannotated 24 hours recordings