Intellectual Property Core Design and Verification for Lossless ECG Data Compression Algorithm

• Main Authors: Wen-Jie Chen, 陳文傑
• Other Authors: 呂志誠
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/3x5ph3

碩士 === 國立臺北科技大學 === 機電整合研究所 === 105 === In recent years, with the increase of the community aging and cardiovascular disease patients, personal home care and telemedicine care are the future trend of development. At present, there are many ECG monitoring patch for the public to use. In order to record and transmit the long-term ECG signals from patients, ECG signal compression has become an important issue. However, in addition to reducing the amount of data compression system, the key issue is to retain the ECG feature waveform characteristics for diagnosis. This paper proposes an ECG compression algorithm for long-term wearable ECG monitoring equipment, which is lossless, low latency, simple operation and low power dissipation. The main design concept is based on the distribution characteristics of ECG adjacent sampling. After the bit separation procedure, the repetitive data will appear in large numbers, and the repetitive data will be encoded to achieve the lossless compression. For the algorithm validation, the MIT / BIH ECG database and the read electrocardiogram test were carried out to discuss the effect of compression rate on different capture frequencies. In the hardware implementation process, the algorithm is implemented on the Xilinx FPGA ML505 to verify the correctness of its function. Finally, the chip is implemented by the TSMC 0.18um Mixed-Signal 1P6M Cell-Base Flow process.