Output details

How many neurons can we see with current spike sorting algorithms?

Spike sorting is one of the main techniques of electrophysiologists to extract maximum information from their recorded data. In this work, we show limitations of spike sorting algorithms that can explain why the number of recorded neurons is much lower than expected from anatomical data, usually attributed to the very low firing of most neurons during recording. The limitations uncovered affected especially the firing of sparse neurons, particularly important for the understanding of memory systems. Based on this work, a new simulation tool was developed in our laboratory to test and improve future algorithms.