Original notes on the kinds of analyses performed by group members:
- Common aspects
- Often need to extract subsets of data, e.g. just some of the trials, ROIs, time, etc.
- General improvements to code reliability & performance possible.
- Antoine
- Various options for averaging, smoothing, normalisation
- Can configure which time region to use for baseline
- Can show volume & ribbon scans in 2d (projection) or 3d (original coords)
- GUI will link between different types of plots when clicked
- Fred
- Smooths speed data in time using Matlab ‘smooth’ function
- Finds where speed above threshold => locomotion
- Baseline: Determines the median grey value for each trace by fitting a kernel function
- Or from the 20% percentile if flag set in some cases
- Sometimes subtracts background before calculating
- Groups ROIs into axons and smooths in time (with some end effects?)
- Time & df/f for axon is mean of the constituent ROIs
- Plotting alongside speed data; overlays with detected locomotion periods
- Trying to detect bursts, and when they start wrt locomotion & puff
- Using an event detection function from 1997 - experimental? Would be better to use Neuromatic for this kind of thing really.
- Hana & Sadra
- Computing dF/F, with baseline taken from mean of initial period
- Correlations between ROIs, dF/F against speed data - statistics & plots, per trial & averaged over trials
- Plotting locations of ROIs on raw image
- Looking at video analysis using third party codes, but still early stages - whisker angles over time, paw location heat map, etc.
Original notes on the kinds of analyses performed by group members: