Electromyogram data compression using single-tree and modified zero-tree wavelet encoding

Abstract

The long-term analysis of the neuromuscular systems, and applications in telemedicine, make electromyogram (EMG) data compression a subject of great practical importance. However, in spite of the increasing demand, only a few studies have been published on this subject. In this paper, we present two wavelet-based lossy compression techniques for EMG data. We propose modifications to the so-called 'embedded zero-tree wavelet coder', which yield very good results in ECG compression applications. We have implemented the algorithms in Matlab and C++ and tested then with several EMG recordings.