ECG Data Compression Through Adaptive Sampling And Arithmetic Coding
G. Lifiana, Martinez Fernando, P. Gianfranco · 2005
ECG data compression is necessary in many applications. Especially in continuously storing 24-hour ambulatory ECG's. There is a current interest in developing new methods and algorithms that can be used in the design of digital H olter devices. This paper presents a novel approach to this problem. ECG data compression is achieved in two stages. The first performs compression by adaptive sampling using, both, the Turning Point and FAN (SAPA) algorithms. The second part is an implementation of a modified arithmetic coding algorithm that does not introduces changes in the data, being completely reversible. Using a test sample composed of 126 extracts of 20 secs, 8-bit resolution, selected from the MIT/BIH arrhythmia database, an average compression of 1 1.29: 1 has been obtained.