Smooth White Matter and Grey Matter contours in 2D visualization

[neuroBazzi]

Client is asking to increase the resolution of the 2D slices and 3D volumes to avoid 'pixelated' effect. I am assuming this would come from @stephenlenzi first, then we would upload it in the app @afonsobspinto

[zsinnema]

@neuroBazzi correct, if the resolution if upscaled then it will take about 2h (including testing) for @afonsobspinto

[afonsobspinto]

We need to have the new upscaled images for grey and white matter, canal, grid and laminas (contour, dashed and filled) We have django management commands to generate those (generate_annotation_images.py, generate_canal_images.py, generate_grid_images.py, generate_lamina_images.py), and what I tried to do was to adapt those according to a set of scripts @stephenlenzi shared with me: scripts.zip

This issue is still pending because currently we are facing an issue, most likely caused by the upscaling of the atlas, that made the canal data disappear. @stephenlenzi already got informed about the problem.

Branch feature/243 contains the latest developments for this issue. (not a typo, I did it alongside my work for #243) I added a couple of comments and refactored the code a bit to make it easier for new eyes to understand it. Be aware that I did not had the chance to properly test the refactored code.

[stephenlenzi]

I believe I have figured out the source of the problem, there was indeed a bug in the upscaling function that meant that the canal would not be present on a subset of annotated images after upscaling. This bug would have affected some of the early manually annotated images (C1-4) but also seems to have affected all of the images with only 3 labels. The solution is a minor fix and i'll push the changes asap

[stephenlenzi]

The fix is on the contours branch see this commit for the updated code

[neuroBazzi]

@afonsobspinto check this out