A compression of ECG data based on neural network

ECG data compression is one of the most important subjects arising in biosignal processing. If a backpropagation (BP) artificial neural network is adopted to compress ECG data, the compression precision speed is low. The system we established improves the speed. The system uses four filters to process ECG data for improving the quality of ECG signals, and adopts the improved first difference algorithm to detect R points accurately and easily. This processing improves the following compression. During the compression, we combine the BP algorithm with the TP algorithm and establish a weight template library to improve the rate of compression. The compression precision and compression ratio can meet the requirements and the processing of ECG data containing a few types of abnormal ECG waves in real time with the system