An Improved ASCII Character Encoding Method for Lossless ECG Compression

Storage and transmission limitations have made electrocardiogram (ECG) data compression an important aspect for ECG computerized systems. In this paper a lossless method based on modified American standard code for information Interchange (ASCII) character encoding for ECG data compression have been proposed. The Proposed method consists of compression algorithm comprising sign count; generation of array representing ECG sample’s signs (+ve, -ve alternatively), adaptive amplification factor; and grouping method of ECG samples and a reverse process for ECG reconstruction. The ability of the proposed compression algorithm has been investigated on the MIT-BIH Arrhythmia Database. The average percentage root mean square difference (PRD) of about 0.32, compression ratio (CR) of about 8.38, quality score (QS) of about 26.02, percent root mean square difference normalized (PRDN) of about 0.56, root mean square error (RMS) of about 0.0018 and SNR of about 45.46 was achieved on MIT-BIH data. The method is also compared with other compression algorithms and showed superior performance in term of PRD, CR, QS, PRDN, RMS and SNR. The novelty of proposed method is the nearly exact reproduction of the original signal (PRD=0.32) and a moderate CR.