自适应噪声减法器用于提高体感诱发电位信噪比

Application of adaptive noise canceller for improving the signal/noise ratio of somatosensory evoked potential

目的利用自适应方法实现从强噪声背景中快速提取体感诱发电位信号。方法采用最小均方误差算法对自适应噪声减法器进行设计，以15例接受脊柱外科手术的病人为对象，将600次刺激响应的SEP做叠加平均得到的信号作为标准的SEP模板，用脑电图（EEG）加白噪声来模拟临床中干扰SEP的噪声。通过仿真实验检验自适应噪声减法器提取SEP信号的性能。结果自适应噪声减法器可以将信噪比提高25dB左右：自适应噪声减法器提取出的SEP信号与SEP模板信号间潜伏期和峰峰幅值的百分比误差均分别小于1％和6％。结论运用白适应噪声减法器可以实现从强噪声背景中快速提取SEP，在提高检测信号信噪比的同时。更快地识别SEP的潜伏期及幅值。

Objective To quickly extract the somatosensory evoked potential （SEP） signals from strong noisy background with adaptive technique.Methods The adaptive noise canceller （ANC） was designed based on least mean square （LMS） algorithm. Fifteen patients during spinal surgery were enrolled in the experiment. The adaptive noise canceller （ANC） was adopted to evaluate its potential use in SEP measurement and a simulation study was conducted on these patients. The 600 times SEP signal was regarded as standardized template signal of SEP while electroencephalogram （EEG） and white noise were regarded as the noise interfering SEP in clinical operations. Results The ANC can elevate the signal/noise ratio at least 25 dB. The errors of latency and amplitude between the results of ANC and the template of SEP signal were less than 1% and 6 % respectively. Conclusion It is demonstrated that the application of ANC can quickly extract SEP signals from strong noisy background and significantly improve the signal/noise ratio. It would be helpful in more quickly measurement of SEP latency and amplitude.