Adaptive data compression for a digital Holter monitoring system

Abstract

Building entirely digital Holter monitoring systems is now pursued and development of digital electrocardiographs is progressing. Conventional Holter electrocardiographs are devices that record electrocardiograms on magnetic tape in the form of an analog signal, which implies a number of problems related to accuracy (waveform distortion) and vibration of movable parts (such as motors), as well as weight and size of the device. This paper discusses a simple Holter electrocardiograph that converts 24-hour electrocardiograms into digital signals, with provision for noise suppression, elimination of baseline fluctuation, easy diagnostics, storing data on media, and data transfer. An entirely digital design for such a device would contribute to circuit integration and would solve the problem of miniaturization by using IC memory cards as the media. However, to record a digital ECG signal for 24 hours, 64.8 MBytes of memory is required for the standard pattern of 2 channels, 12 bits, and 250-Hz sampling. IC memory cards currently available lack this capacity, and an algorithm for data compression is necessary. In this paper, a simple algorithm is proposed for R-R interval extraction in a compact Holter monitoring system, as well as a data compression algorithm that provides adaptive regulation of compression ratio and accuracy. Simulation results are reported.