ECG DATA REDUCTION FOR MEDICAL TELEMETRY SYSTEM

Masaki Kyoso, Akihiko Uchiyama
School of Science and Engineering, Waseda University
3-4-1 Ohkubo Shinjuku-ku, Tokyo, 169, Japan

INTRODUCTION Many ECG data reduction methods had already been developed in the past twenty years. Although these methods have acceptable compression performance, the total characteristics such as channel noise and power consumption of transmitters are not considered. Since percentage of abnormal waveforms are not so large in case of ECG monitoring, transmitter radiates continuously even if ECG is normal. This leads to signal collision between transmitters and also battery consumption.

METHODS The data reduction methods we have presented was designed for wireless ECG telemetry system. The transmitter has a simple ECG analyzer which recognizes waveforms are normal or abnormal. If abnormal ECG is detected, full ECG waveforms are compressed with several waveforms prior to detected abnormal ECG. An important parts of the ECG waveforms, such as QRS complex, P wave, T wave are checked and coded with error correcting code. During normal ECG is recognized, data reduction (R-R interval or only heart rate per 30 seconds) is performed. The compression ratio is far better than the other methods, because this algorithm rejects redundancy of data by means of the knowledge of percentage of normal ECG.

RESULTS The advantages of this method are high channel efficiency, low power consumption, because the percentage of normal data is sufficiently larger than that of full coded transmission. A data reduction of more than expected were obtained in the experiments in both cases of ECG from generator and patient.

DISCUSSION We have evaluated the performance of proposed system by computer simulation and experiments. Comparison of transmitted and received waveforms of abnormal case showed enough quality for ECG monitor. We can conclude very high compression ratio can be obtained by this method. However, the practical system evaluation is required in hospitals.