Evaluation of the `TRIM´ ECG data compressor

Data compression for storage of long-term ECGs is increasingly common, due to the recent introduction of devices which perform the functions of Holter ECG recorders using solid-state memory rather than tape. Often, the quality of compressed ECGs obtained from these devices is poor. The authors present a method for evaluating ECG data compressors for clinical acceptibility and a case study in which they evaluate the turning-point recursive-improvement (TRIM) compressor developed by G.B. Moody et al. (1987). It is concluded that TRIM, when adjusted to produce a mean output rate of 200 bits per channel per second, does not measurably influence the accuracy of clinicians´ diagnoses with respect to cardiac rhythm or conduction. The evaluation methodology is general and suitable for standardized testing of ECG data compressors