Wavelet-Fuzzy Based High-Accuracy R-wave Identification for Electrocardiography (ECG) Bio-Signal Processing and Prototyping System/FPGA Realization

• Main Authors: Yang-Zhe Lin, 林陽哲
• Other Authors: Yu-Cherng Hung
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/59998818192619540122

碩士 === 國立勤益科技大學 === 電子工程系 === 100 === Electrocardiogram (ECG) information in general can be used for the detection of human heart disease, such as detections of arrhythmia, myocardial infarction, ventricular hypertrophy and ventricular fibrillation. A large amount of data is needed to read before making a correct decision. The exactly identification rate depends on the knowledge-and-experience of doctors. As a result, automatic identification of ECG signal by using a test equipment to help doctor effectively making a good judgement is necessary. QRS waveforms are the major components in ECG signal. In this research, there are three relative ECG signal processing subjects. First, we improve a algorithm complexity for QRS waveform identification. Comparing with reference [1], the overall executive time of the improved algorithm is reduced as 120 ms, faster than [1] in the same identification rate. In the second research, based on wavelet transform and fuzzy logic, a new algorithm for highly precision R-waveform identification is proposed. The test data of arrhythmia is provided by Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH). The test data are total 48 recorders and can be obtained in the academic open resources. After software simulation, the new algorithm sucessfully achieves 99.87% identification rate. In addition, the function of wavelet transform is realized by FPGA chip of Altera company. Finally, in the third research, a prototyping system for Bio-ECG signal capturing and QRS identification is sucessfully realized. The hardware includes self-design PCB board for signal amplicication and a data acquisition module of National Instrument company. The software are adopted by LabVIEW and Matlab to achieve QRS waveform identification. Afterwards the prototyping system can be expanded more functions for various applications of ECG signal processing. The system is expected to on-line help doctor quickly making a judgment on various heart diseases.