定点数自适应消噪器用于体感诱发电位检测  

Application of Fixed-point-based Adaptive Noise Canceller for Somatosensory Evoked Potential Detection

作　　者：张宇奇[1] 沈冲飞[1] 谢小波[1] 崔红岩[1] 陆瓞骥[2] 胡勇[1,2] 

机构地区：[1]中国医学科学院生物医学工程研究所,天津300192 [2]香港大学矫形与创伤外科学系

出　　处：《医疗卫生装备》2009年第5期29-31,共3页Chinese Medical Equipment Journal

摘　　要：目的:为了改善脊髓术中监护的准确性和实时性,一种可移植于FPGA的定点数算法自适应噪声消除器被应用于改善体感诱发电位的信噪比。方法:仿真浮点数算法、定点数算法自适应消噪器对离线SEP信号处理的效果。比较两种滤波器在不同收敛系数下失真指数的趋势。结果:定点数算法在收敛系数较小的范围内失真指数较浮点数算法略大。随着收敛系数的增大,二者差异逐渐缩小。结论:通过合理优化收敛系数,定点数自适应消噪器能够逼近浮点数自适应消噪器,满足SEP检测的要求。Objective In order to improve the accuracy and real-time of intra-operative monitoring, a fixed-point-based adaptive noise canceller（ANC） is used in improving signal to signal-noise ratio （SNR） of Somatosensory Evoke Potential（SEP） extraction. Methods The effect of signal processing in simulation of floating-point algorithm and fixed-point algorithm with ANC to SEP offline browse, the trend of distortion index under different coefficient convergences（ ）were compared in two filters. Results The distortion coefficient is slightly larger than the fixed-point algorithm in the smaller extent, while the increasing convergence of the factors is differently gradually narrowed. Conclusion By optimizing the value selection, fixed-point algorithm ANC can get as good results as floating-point algorithm.

关 键 词：体感诱发电位 自适应噪声消除 现场可编程门阵列 定点数算法 收敛系数 信噪比 

分 类 号：R318.6[医药卫生—生物医学工程]