Armand Collin

Update: yesterday I bought a used iMac so I'll definitely be able to help with this in the following weeks. I would start with what @Stoyan-I-A drafted and then create...

oops forgot this one: ## 5) "Onion bulb" myelin This pathology can only be observed on one data-point. The features of this sample look very different (contrast/content-wise) from the rest...

I think it would be clearer if we moved the "Example dataset" section under the "Segmentation - syntax" section so that the examples and the link to the data are...

Now that the ivadomed integration is done, models can now be trained using anisotropic pixel sizes (resolution/patch size). However, in #620, this discussion was raised again because anisotropic pixel sizes...

We just found out about the [empanada](https://github.com/volume-em/empanada) project which leverages `napari` for their GUI: https://github.com/napari/napari As discussed in #544, maybe this would be a good migration target for the FSLeyes...

Note regarding this project: there is a similar problem with a "multi-contrast pipeline", which is that we need to encode the metadata as input data so that the model knows...

Output on master: ``` » axondeepseg -s 0.211 -i 22G-13-33_20x.tif -m /home/herman/Documents/NEUROPOLY_21/crypte/20220316_upload_wakehealth_model_on_gh/model_seg_human_axon-myelin_bf -t BF -z 0.7 2022-07-27 08:39:28.751 | INFO | AxonDeepSeg.segment:main:323 - AxonDeepSeg v.4.0.0 Lossy conversion from float32 to...

With this quick fix, the only remaining warnings are those: ``` Lossy conversion from float64 to uint8. Range [0, 1]. Convert image to uint8 prior to saving to suppress this...

> As a first example of potential breaks, forcing all saved images to 8bit won't be compatible with this new feature you re interested in implementing #668, so maybe it...

> On the IVADOMED side, @mariehbourget mentions it here: [#175 (comment)](https://github.com/axondeepseg/axondeepseg/issues/175#issuecomment-1178042591) > > That it's likely this function: > > https://github.com/ivadomed/ivadomed/blob/a58898448b69dddfdccffd8de1694a70d78c900b/ivadomed/inference.py#L215-L228 > > I don't know if a solution like...