Mathieu Boudreau, PhD

Improve and force imread/imwrite conversion to 8bit int

27

## Checklist - [x] I've given this PR a concise, self-descriptive, and meaningful title - [x] I've linked relevant issues in the PR body - [x] I've applied [the relevant...

Resolve warning during segmentation about lossy conversion (int64 to uint8)

12

This message (I believe) appears during any segmentation attempt (i.e. Jupyter Notebooks/Python env/command line). The images segment fine, so to resolve this warning message it's just a matter of doing...

Suppress warnings

1

**branch**: mb/suppress-warnings related issues: #175 We've accumulated a lot of warnings, so I'll go through them here and find solutions to suppress as many of them as I can. Here's...

From this discussion: https://github.com/axondeepseg/axondeepseg/pull/629#discussion_r914951056 It should be handled in the code, and maybe send a user-friendly alert in the GUI.

Add feature to filter axons in segmented images using a list of axon indices

2

Example workflow: * Segment an image * Create morphometrics file * Identify axons in morphometrics file that have unlikely properties (eg very small or large g-ratios, single pixel axons, etc)...

See this discussion: https://github.com/axondeepseg/axondeepseg/discussions/652

Change demo image in documentation

7

Throughout the documentation, we give examples using `test_segmentation/test_sem_image/image1_sem/77.png`, however I don't think this image exists anymore. I think we should instead change it to `AxonDeepSeg/models/model_seg_rat_axon-myelin_sem/data_test/image.png`, which is the SEM test...

Is axondeepseg compatible with anisotropic pixel sizes, and if not how to we do so?

4

e.g. * Patching/training? * segmentation * calculating morphometrics

New feature: pixel units

2

Add the option to set pixel units. Should be available on the command line and FSLeyes plugin.

Create Docker/Singularity container for training ADS models

1

See this project board https://github.com/ivadomed/ivadomed/projects/8#card-77637238 for some current work on the IVADOMED side.