Project-MONAI / Project-MONAI/MONAILabel

Issues with multiple labels and masks overlapping in Monailabel

Open
#1,856 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
891
Forks
269
Avg merge
15h 41m
Merged PRs (30d)
1

Description

Hey everyone, I’m labeling brain segments in MONAILabel.

The masks / ROIs used as input for Monailabel are a subselection extracted by slicerio.extract_segments() function from a 3DSlicer Segmentation node.

I modified the config files from the models DeepEdit and Segmentation (from the radiology app) to match our masks, but I haven’t made much progress so far.

Here is a screenshot of the DeepEdit config file

Image

I’ve included 5 masks so far, which are clearly identified in 3D Slicer (see figure below):

In 3D Slicer, the 5 labels are separate and clearly identified

Image

However, when I open them in MONAI Label, the labels often get collapsed. For example, the ipsilateral and contralateral hemisphere labels may collapse into the contralateral hemisphere, the cerebellum and lateral ventricles become one label, and some labels are completely absent (see figure below):

In Monailabel, all 5 labels are collapsed into 3 - cerebellum, ipsilateral and contralateral lateral ventricles

Image

Has anyone encountered this issue before or have any advice on how to fix it?

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the reported slicerio.extract_segments() output and the modified DeepEdit and Segmentation configuration files, then reproduce the label-loading behavior in MONAILabel with the five masks. Done means identifying why distinct labels are collapsed or omitted and confirming that all five labels remain separate when opened in MONAILabel.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.