The Characteristic Extraction And Data Compression Of ECG Signal Base On Wavelet Transform

Cardiac diseases are one of the major diseases that harm human health. With the further development of the analysis of ECG, the detection of ECG is one of the most important methods for clinical diagnosis of cardiac diseases as QRS waves in electrocardiogram contain abundance information about person heart. The research and detection of QRS can improve the performance of the unusual heart rate guardianship and the Holter ECG analysis system. The detecting of QRS wave is always the research focus in which some concerned questions are how to detect the R wave in accuracy, how to detect the feature and compress the data for saving and transmitting. This paper is just trying to resolve the question, using wavelet theory to research feature detecting and extraction and the data compression of the ECG signals.In this paper, after studying the knowledge about Wavelet Transform (WT) and base on the predecessor's achievement, first analysis and discuss the related knowledge of ECG, recommend the basic conception, application and actuality about the wavelet analysis theory. By comparing and analysis several wavelets used in the signal processing, combining the characteristic of ECG signal choose a proper wavelet. Then we used the wavelet theoretic to make some examination for the existing ECG signals and mostly aim at the noise elimination, the R wave detection, characteristic detector and the data compression, then analysis and elicit some effective method. And use this method to process and analysis the ECG signals in the MIT/BIH databases. The method that used in this paper was examined by the testing data and the examination indicated that by this method the real-time, anti-jamming and the accuracy of the detecting has a great improvement.Based on the analysis of all kinds of data compression algorithms combining the characteristic of ECG signals, we choose the algorithms which combined the difference compression algorithm and Discrete Cosine Transform (DCT) compression algorithm to compress and reconstruct the ECG signals. The result shows that the reconstructed signals have a high fidelity, the wave is mild and has a high compression rate. Meanwhile the algorithm is so simple and normative, and the realize cost is inexpensive, that it has a great use value.