Two-Dimensional Compression of ECG Signals Using HEVC-Intra Encoder and Pre-processing Techniques

Abstract

The paper presents the proposal of an electrocardiogram (ECG) compressor based on two-dimensional coding. The signal is pre-processed and the R waves detected, from these waves the signal is segmented and arranged in an N × M matrix, composed of M segments and N the maximum length between two R waves. Once this matrix is composed, the interpolation process is performed so that all segments M have the same size N. With the interpolated matrix a pre-processing is applied to improve the two-dimensional correlation of the matrix. This is done by rearranging the segments so that the correlation of the segments be maximized. Some widely-adopted image compression algorithm is used to reduce the volume of data. In the present work, the intra mode of the High Efficiency Video Coding (HEVC) is applied to encode real ECG signals obtained through the MIT-BIH Arrhythmia Database. These signals impose a greater challenge for the proposed system. The application of this proposal demonstrates that the use of HEVC-intra image encoder is efficient in ECG compression. A quantitative performance evaluation was performed and compared to other results found in the literature.