中国物理B ›› 2006, Vol. 15 ›› Issue (1): 100-103.doi: 10.1088/1009-1963/15/1/016
朱学敏, 任育峰, 于洪伟, 赵士平, 陈赓华, 张利华, 杨乾声
Zhu Xue-Min (朱学敏), Ren Yu-Feng (任育峰), Yu Hong-Wei (于洪伟), Zhao Shi-Ping (赵士平), Chen Geng-Hua (陈赓华), Zhang Li-Hua (张利华), Yang Qian-Sheng (杨乾声)
摘要: Considering the properties of slow change and quasi-periodicity of magnetocardiography (MCG) signal, we use an integrated technique of adaptive and low-pass filtering in dealing with two-channel MCG data measured by high $T_{\rm c}$ SQUIDs. The adaptive filter in the time domain is based on a noise feedback normalized least-mean-square (NLMS) algorithm, and the low-pass filter with a cutoff at 100Hz in the frequency domain characterized by Gaussian functions is combined with a notch at the power line frequency. In this way, both relevant and irrelevant noises in original MCG data are largely eliminated. The method may also be useful for other slowly varying quasi-periodical signals.
中图分类号: (Biophysical techniques (research methods))