What is the aim of the package ?

[NathanMolinier]

From what I understand, the goal of this package is to create a slicer module to easily segment pathologies/anatomies on medical images (CT, MRI and more ?).

However, when I look at the code, it seems to rely heavily on multiple configuration files (.yml) that are not completely intuitive and its hard to know to what extent these configuration files are customizable:

• classification_config.yml
• general_config.yml
• keyboard_shortcuts_config.yml
• label_config.yml

So what I don't get is this:

1. Do we want to allow the user to create/add buttons or short cuts only by modifying the configuration files ? If so, how to do it ? What are the limitations ?
2. Do we want to provide the user with determined configuration files that correspond to different workflows ? (e.g. MRI lesion segmentation and classification, MRI rootlets segmentation, CT edema segmentation...)
3. Why do we have that many configuration files ? Can't we regroup ?

I personally think that these questions need to have a precise answer to clarify the direction of the project. I'm no expert with slicer but currently all the possibilities feel really overwhelming.

[jcohenadad]

Agreed-- it comes back to the 'big picture' discussion that we should have as a team, tomorrow and in subsequent meetings.

[maxradx]

Will be directly addressed deeply in the updated README.md from issue 62. Note that the specific aim of the package is to optimize worklow related to manual annotation, classification and visualization of medical imaging volumes.

[maxradx]

@NathanMolinier
From what I understand, the goal of this package is to create a slicer module to easily segment pathologies/anatomies on medical images (CT, MRI and more ?). --- see the Purpose section in PR 105 (#105)

Do we want to allow the user to create/add buttons or short cuts only by modifying the configuration files ? If so, how to do it ? What are the limitations ? --- yes we want to user to specify basic functions accordingly to its preference (ex keyboard shortcuts) That should be able to be done from the GUI configuration window.

Do we want to provide the user with determined configuration files that correspond to different workflows ? (e.g. MRI lesion segmentation and classification, MRI rootlets segmentation, CT edema segmentation...) --- yes that is something that still remains to be implemented, but for which work is in progress.

Why do we have that many configuration files ? Can't we regroup ? --- done in previous PR (PR 97 #97)

In brief: issue now outdated. Specific questioning addressed in PR 105, so please refer to it if anything still unclear.

Thanks.