Editor’s Pick: A much needed comparison of spike sorting algorithms.

Reviewing Editor Arvind Kumar, Ph.D. selected this paper and explains why he considers it noteworthy.

Spike sorting is an essential step before we can start to interpret the extracellularly recorded spiking activity. With an increase in high-density electrode arrays such as neuropixels spike sorting has become even more crucial. While a number of spike sorting tools are available and used, surprisingly there is no systematic comparison of the various spike sorting algorithms using data for which ground truth is known. This paper fills this knowledge gap.

In particular, authors have discussed the effect of spatially and temporally synchronous events (the so-called spike collisions). They show that the intuitive template matching algorithms are more accurate than more sophisticated algorithms. But all spike sorting algorithms are affected by synchronous events and this should be kept in mind when estimating and interpreting spike time synchrony. Somewhat reassuringly, the performance of most of the algorithms was not affected by the firing rate and synchrony in the neural activity.

Overall, I think this paper is a valuable contribution and should guide neuroscientists in choosing an appropriate spike sorting algorithm and I hope it will also inspire further improvements in automatic spike sorting algorithms.

Read the full article:

How do spike collisions affect spike sorting performance?
Samuel Garcia, Alessio Buccino, and Pierre Yger