search

cortex-lab / KiloSort

GPU code for spike sorting
GNU General Public License v2.0
175 stars 100 forks source link

biased against narrow wave-shape spikes? #235

Closed esiabri closed 3 years ago

esiabri commented 3 years ago

Hi, thank you for this amazing tool! I've noticed that it seems that the narrow wave-shape spikes (coming from putative FS neurons) are more probable to be missed. I have attached here two sample snapshots from the Kilosort GUI, as it can be seen spikes with lower amplitudes were deleted while larger narrower spikes are missed. I was wondering if I miss something and I can fix this through some parameters? does it have something to do with the fact they are less probable to be detected on multiple channels? Capture1 Capture2

esiabri commented 3 years ago

this was with kilosort 3 so I'll move it to the kilosort 3 repository!

cgallimore25 commented 3 weeks ago

this was with kilosort 3 so I'll move it to the kilosort 3 repository!

@esiabri I'm finding this same issue in my recordings using KS3. i know this is an old post, but did you ever figure out a combination of parameters that enable KS to detect these spikes?