bug:Issue with running FMCIB negative controls with parallel jobs on H4H

[JackieC13462]

Summary: Negative control analysis fails for certain permutations and regions in H4H, resulting in missing output data.

Description: When running negative controls in H4H using the command pixi run readii_negative DATASET, data for specific permutations and regions is missing in the output for multiple samples. Suspect this is related to the parallel flag when using the "non_roi" region. When parallel is set to false, the controls generate completely, but when set to true, some controls aren't generated.

Negative Control Parameters:

• parameters are written into the config file for the given dataset
• IMAGE_TYPES:
  • regions: "full", "roi", "non_roi"
  • permutations: "randomized", "sampled", "shuffled"
  • crop: "cube"
  • resize: ``

Reproduction:

1. Run the command: pixi run readii_negative DATASET with the above parameters.
2. Observe missing data for different permutations and regions.
3. Check readii index that's generated with the negative controls

Additional Information:

• The CLI command executes without errors.
• The entire process completes successfully, despite the missing data.