Playing with some calcium imaging data from Marina.
First thing first, motion correction with ecc! Looks like it works pretty well. I downsampled the original data to 256x256x600 (originally 512x512x600), took about 10 minutes to run the motion correction on my desktop.
Here are low res movies showing the motion correction.
Before motion correction:
The flashing of the movies is just an artifact because I did not normalize all the frames together.
Here is the motion corrected data in mat format.
Ok, so here are all the cells I can see with ICM in this data set:
This was run using PCs 1:50. The rest of the ICs are all noise and edge artifacts.
And here are the ROIs that are generated for these cells: