Challenges for automated spike sorting: beware of pharmacological manipulations

TL;DR

This paper highlights the challenges in automated spike sorting, emphasizing the impact of pharmacological manipulations on spike waveforms and the need for algorithms to account for biological variability.

Contribution

It underscores the importance of considering drug-induced waveform changes in developing robust automated spike sorting algorithms.

Findings

Spike waveforms are affected by pharmacological manipulations.

Current algorithms may fail under biological changes.

Future methods should incorporate biological constraints.

Abstract

The advent of large-scale and high-density extracellular recording devices allows simultaneous recording from thousands of neurons. However, the complexity and size of the data makes it mandatory to develop robust algorithms for fully automated spike sorting. Here it is shown that limitations imposed by biological constraints such as changes in spike waveforms induced under different drug regimes should be carefully taken into consideration in future developments.