关于手动尖峰分类的变异性

On the variability of manual spike sorting.

作者信息

Wood Frank, Black Michael J, Vargas-Irwin Carlos, Fellows Matthew, Donoghue John P

机构信息

Department of Computer Science, Brown University, Providence, RI 02912, USA.

出版信息

IEEE Trans Biomed Eng. 2004 Jun;51(6):912-8. doi: 10.1109/TBME.2004.826677.

DOI:10.1109/TBME.2004.826677

PMID:15188858

Abstract

The analysis of action potentials, or "spikes," is central to systems neuroscience research. Spikes are typically identified from raw waveforms manually for off-line analysis or automatically by human-configured algorithms for on-line applications. The variability of manual spike "sorting" is studied and its implications for neural prostheses discussed. Waveforms were recorded using a micro-electrode array and were used to construct a statistically similar synthetic dataset. Results showed wide variability in the number of neurons and spikes detected in real data. Additionally, average error rates of 23% false positive and 30% false negative were found for synthetic data.

摘要

动作电位（即“尖峰信号”）分析是系统神经科学研究的核心内容。尖峰信号通常通过人工从原始波形中识别出来以供离线分析，或者通过人工配置的算法自动识别以供在线应用。本文研究了人工进行尖峰信号“分类”的变异性，并讨论了其对神经假体的影响。使用微电极阵列记录波形，并用于构建统计上相似的合成数据集。结果表明，实际数据中检测到的神经元数量和尖峰信号存在很大的变异性。此外，合成数据的平均误报率为23%，漏报率为30%。