Editor’s Pick: Movement Artifact Correction for Microelectrode Array Recording

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

High-density microelectrode arrays such as Neuropixels probes allow for simultaneous recording of hundreds of neurons. Such probes have rapidly become a popular way to record electrical activity of neurons. However, these probes are susceptible to movement artifacts, which arise due to relative motion of the cortical tissue and the probe. Such artifacts compromise the accuracy of automatic spike sorting algorithms and that of the subsequent data analysis.

In this paper, Nicholas Watters et al. present a new method for motion drift correction for microelectrode arrays. They note that when there is no motion artifact, the distribution of the data in spike amplitudes–neuron depth space would be sparse. Motion artifacts will make the distribution non-sparse. They exploit this observation to build a generative model and learn the motion function to maximize the sparsity of the amplitude–depth distribution.

This is quite an elegant solution, which outperforms the available state-of-the-art algorithm. I think this algorithm will be a very useful resource to facilitate more correct spike sorting of extracellular activity.

Read the full article:

MEDiCINe: Motion Correction for Neural Electrophysiology Recordings
Nicholas Watters, Alessio Buccino, and Mehrdad Jazayeri