Technology-Aware Algorithm Design for Neural Spike Detection, Feature Extraction, and Dimensionality Reduction
Technology-Aware Algorithm Design for Neural Spike Detection, Feature Extraction, and Dimensionality Reduction
复制标题
DOI：
10.1109/tnsre.2010.2051683
复制
发表时间：
2010-10-01
影响因子：
4.9
通讯作者：
Markovic, Dejan
中科院分区：
文献类型：
--
作者：
Gibson, Sarah;Judy, Jack W.;Markovic, Dejan
诸如脑机接口之类的应用需要硬件尖峰排序，以便1）获得单个单元活动，以及2）执行用于无线数据传输的数据简化。这样的系统必须是低功率、低面积、高精度、自动化的，并且能够在真实的时间内操作。几个检测，特征提取和降维算法的尖峰排序的准确性与复杂性进行了描述和评估。选择非线性能量算子作为最佳的棘波检测算法，它对噪声具有最强的鲁棒性，并且相对简单。离散导数被选为最佳的特征提取方法，保持高精度的信号噪声比的复杂性的数量级小于传统的方法，如主成分分析。我们介绍的最大差异算法，这是最好的降维方法，硬件尖峰排序。
Applications such as brain-machine interfaces require hardware spike sorting in order to 1) obtain single-unit activity and 2) perform data reduction for wireless data transmission. Such systems must be low-power, low-area, high-accuracy, automatic, and able to operate in real time. Several detection, feature-extraction, and dimensionality-reduction algorithms for spike sorting are described and evaluated in terms of accuracy versus complexity. The nonlinear energy operator is chosen as the optimal spike-detection algorithm, being most robust over noise and relatively simple. Discrete derivatives is chosen as the optimal feature-extraction method, maintaining high accuracy across signal-to-noise ratios with a complexity orders of magnitude less than that of traditional methods such as principal-component analysis. We introduce the maximum-difference algorithm, which is shown to be the best dimensionality-reduction method for hardware spike sorting.