HELP: Spike sorting for double sided probes

[bd125]

Hi! I am using a double sided 128 channel probe from cambridge neurotech (ASSY-325D H7). It is a double sided probe, but I don't know how to corporate this probe map into kilosort, since kilosort only takes the xy coordinates of the probe, right? Any suggestions? Thank you so much!

[jacobpennington]

Correct, Kilosort4 does not support z-axis information. The first thing I would try is simply duplicating the contact positions but shifted over by the shank width (60um) on the x-axis. That should allow spikes detected on opposite sides to be clustered together, if some of the waveforms are being picked up on both.

If you don't expect a lot of overlap in activity on both sides, or if you see some strange behavior with that, then you can just space them apart by some arbitrary large amount (like 200um) and treat them as separate shanks. Any units that show up on both "shanks" can still be merged together after clustering.

[bd125]

Thanks for the suggestion! I will give a try and post an update later

[jacobpennington]

Closing this for now, please feel free to re-open it later when you have an update.