Channel Map Creation

[nandi-nilanjana]

Hi, I am using recordings from custom-made chronic laminar implants. Since no one has used such implants, I cannot verify my channel map template with anyone else.

I have probes which have two shanks with 16 contact points per shank. Since these are chronic , there is no drift issue or overlapping since these probes are placed on particular sites. So I am only giving the depth (ycoords) of each of the contact points for the channel map. I have 456 channels in total (448 neuronal recoding channel + 8 analog).

So the depths go from 0 to -3750 with 150 micron spacing for each shank. This depth variation multiple by 8 shanks (from my 4 probes) would be the full ycoords in my channel map. I am able to run kilosort and visualise in phy. But I am not sure if this is the correct way of giving channel map.

Example, if I see single unit activity (SUA) on a particular channel during recording, after using kilosort I detect the SUA in an adjacent channel. Is this an issue with the channel map ?

Looking forward to getting back some feedback.
Thank you!

[jacobpennington]

Can you please upload kilosort4.log from the results directory, or add a screenshot of the probe preview from the GUI?

[nandi-nilanjana]

Hi, I think I might have gotten the issue with detecting spikes on a different channel number. I realized its due to the indexing in python that starts from 0. So the spikes which I see in a particular channel while recording would be detected in the (channel - 1) index in phy.

But I am still not confident about my channel map. I am using kilosort2, but I only see the phy.log file in my output directory not a kilosort.log file.

I am attaching the screenshot of probe view in phy and kilosort (before running) from an example session-

The problem is I don’t see the correct depth in phy based on the ycoords I give in my channel map. For example, let’s say in my channel map, the ycoords of channel 11 has depth of 1000. But in phy, in the depth column (refer to the screenshot), the same channel 10 (index -1 in phy) shows a depth of 0.

So I don’t understand how should I then change my channel map.

I have implants in the motor and the parietal areas and each area has 16 shanks which are custom made probes (not NeuroPixel). These are laminar chronic probes. So the x and k coordinates for the channels in the same shank would be the same. I was denoting number 1,2,...upto 32 to all my 32 shanks. For the ycoords I have the actual depths for each shank. This is what I was putting in my channelMap.

Thank you again in advance.

phy.log

[jacobpennington]

There won't be a kilosort4.log file if you're using Kilosort2. We aren't providing detailed support for previous releases anymore. Have you tried sorting this data with Kilosort4?