Closed davidhildebrand closed 9 months ago
I hadn’t thought about it but it’s a good point! Have you run 3D cellpose on your data at all? I think it would be relatively easy to import masks from 3D cellpose and do extraction/neuropil correction/deconvolution using my pipeline. The alternative is integrating cellpose directly into the pipeline, which is probably not feasible in the short term for me.
On 25 Jan 2024, at 20:33, David Hildebrand @.***> wrote:
⚠ Caution: External sender
Have you considered implementing a 3D cellpose-based anatomical ROI detection approach?
We have noticed with the ribosome-tethered indicators in single plane imaging modes that the anatomical ROI detection often tends to outperform the other suite2p approaches. Since cellpose has 3D support, it seems like it might be possible to implement this in s2p-lbm.
— Reply to this email directly, view it on GitHubhttps://github.com/alihaydaroglu/s2p-lbm/issues/56, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADSYD3QSRGSG235BGSZU6STYQK6PVAVCNFSM6AAAAABCLEF642VHI2DSMVQWIX3LMV43ASLTON2WKOZSGEYDCMJSGQZDMNQ. You are receiving this because you are subscribed to this thread.Message ID: @.***>
It turns out that running 3D cellpose on the data isn't producing very good results at the moment.
More related to data/alignment issues than anything else, so this can be revisited if the situation improves.
Have you considered implementing a 3D cellpose-based anatomical ROI detection approach?
We have noticed with the ribosome-tethered indicators in single plane imaging modes that the anatomical ROI detection often tends to outperform the other suite2p approaches. Since cellpose has 3D support, it seems like it might be possible to implement this in s2p-lbm.