Unsupervised Spike Sorting Based on Discriminative Subspace Learning
This work addresses the problem of spike sorting for neuroscience studies, offering incremental improvements in accuracy and robustness over widely used methods.
The paper tackles unsupervised spike sorting by introducing two algorithms based on discriminative subspace learning, which achieve substantially higher accuracy and better cluster separability in lower-dimensional feature spaces compared to existing methods.
Spike sorting is a fundamental preprocessing step for many neuroscience studies which rely on the analysis of spike trains. In this paper, we present two unsupervised spike sorting algorithms based on discriminative subspace learning. The first algorithm simultaneously learns the discriminative feature subspace and performs clustering. It uses histogram of features in the most discriminative projection to detect the number of neurons. The second algorithm performs hierarchical divisive clustering that learns a discriminative 1-dimensional subspace for clustering in each level of the hierarchy until achieving almost unimodal distribution in the subspace. The algorithms are tested on synthetic and in-vivo data, and are compared against two widely used spike sorting methods. The comparative results demonstrate that our spike sorting methods can achieve substantially higher accuracy in lower dimensional feature space, and they are highly robust to noise. Moreover, they provide significantly better cluster separability in the learned subspace than in the subspace obtained by principal component analysis or wavelet transform.