loading page

Spike sorting using Hidden Markov Models
  • Roger Herikstad,
  • shihcheng
Roger Herikstad
National University of Singapore
Author Profile


Recent years have seen a steady increase in the potential number of neurons that can be recorded simultaneously. Yet, the techniques for extracting single unit spike trains from multi-channel extra-cellular recordings are still being established. Recently \cite{Herbst:2008kn}, introduced a spike sorting algorithm based on fitting a Hidden Markov Model to the joint firing of neurons producing a recorded signal. Here, we implement their model the Julia programming language, and investigate the advantages of this technique over previous algorithms.