摘要
神经元锋电位(spike)的有效检测在神经信号处理中具有十分重要的作用。为了去除检测后spike中混杂的噪声干扰,提出了一种基于相关距离的k—近邻方法,用于识别spike检测中误检的大幅值噪声信号,并利用仿真实验和实测数据对算法的有效性进行了验证。仿真结果显示,当spike的信噪比大干0 dB时,识别正确率可达90%以上。
Spike detection is crucial for investigating the mechanisms of neural information processing in neuron system. In or- der to improve the spike detection accuracy, a new algorithm based on CDkNN （correlation distance based/e nearest neigh bor） is proposed to remove exactly noise events in data. The effectiveness of CDteNN was verified by simulation and real data. The simulation results show that the recognition accuracy of CDkNN can reach above 90 ~ when the signal-to-noise ratio is more than 0 dB.
出处
《中国科技论文》
CAS
北大核心
2013年第1期46-50,共5页
China Sciencepaper
基金
国家自然科学基金资助项目(60971110)
