Large amount of cells dropping during MiniAn first temporal update

[Paul Vander]

Hi all,

I am having some trouble with MiniAn filtering out multiple cells during the first temporal update that appear to have real temporal activity. A few examples of cells that appear to have real temporal activity, but no "spikes" are detected, so they get dropped during the update:

And strangely enough, for some cells it seems to work fine and detect spikes:

If anybody has any idea of what could be causing this please let me know! I've tried tweaking all of the parameters for the temporal update, but nothing seems to make a big difference.

For context: the data comes from 11 10-minute videos of the medial preoptic area collected about ~1 hour apart from each other. Since I never removed the miniscope, I simply concatenated the 11 videos together before starting MiniAn analysis, since the FOV should be the same. There were also numerous frames with horizontal "striping" patterns that I removed manually, but these are <1% of the total frames across all videos. And here's the results of initialization before the CNMF module:

--

Paul Vander

Pronouns: he/him/his

PhD Candidate, AHA fellow

Correa lab | https://sites.lifesci.ucla.edu/ibp-correalab/

Molecular, Cellular & Integrative Physiology PhD Program

University of California, Los Angeles

[Susie Yu Feng]

Hi Paul,

Have you tried lowering the sparse penalty more in temporal update? The 3rd example you shared (the one with spile fitting) looks a bit weird to me as it doesn't look like real calcium decay curve (but not sure if zoom-in will be the same case)

I'm sharing a list of parameters I used in the past for CA1 imaging which worked well, hope it can be somewhat helpful!

best,

Susie

--
You received this message because you are subscribed to the Google Groups "Miniscope" group.
To unsubscribe from this group and stop receiving emails from it, send an email to miniscope+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/miniscope/CABQnEyi7VADgh5SSX8_9U50v-Dy3vYDGnwk2GLf1E9wZMtj5VA%40mail.gmail.com.

[Paul Vander]

Hi Susie,

Thanks for sharing the details of your parameters! Unfortunately, it seems that changing the sparse penalty doesn't solve the issue, and using the sparse penalty from your spreadsheet seems to result in all cells being dropped...

And here's a zoom in on the spike-fitting for the 3rd example. I think it was hard to see in the original screenshot since there are a lot of frames (100,000+) in the video

If you or anybody else has any other ideas/suggestions they would be greatly appreciated!

[Paul Vander]

Hi all,

Just wanted to follow up that I believe I identified the issue. I used white (not totally opaque) dental cement for the baseplating, and I believe that there may have been some light leakage causing rapid shifts in the fluorescence values in sessions where overhead lights were on in the vivarium. When I removed frames that were collected when vivarium lights were on, the temporal update worked as expected and didn't drop any units.