A novel approach using WAAVES coder for the EEG signal compression
Imen Dhif, Mohammed Shaaban Ibraheem, Laurent A. Lambert, Khalil Hachicha, Andréa Pinna, Sylvain Hochberg, Imen Mhedhbi, Patrick F. Garda · 2016
In this paper, we propose a new approach to compress Electroencephalogram (EEG) signals using the WAAVES compression algorithm and Independent Component Analyses (ICA). Firstly, ICA is applied to the 1D-EEG signals as a preprocessing stage to uncorrelate signals. Then, the output of the ICA is scaled and reformatted into a 2D-matrix to be compressed as an image using the WAAVES coder. This scheme gives a better compression ratio and a better Percentage Root-Mean-Squared Difference (PRD). Our work increases the compression efficiency (CR = 36.60) while reserving the EEG signal diagnostic quality (PRD=4.73).