ENHANCE-PET / MOOSE

MOOSE (Multi-organ objective segmentation) a data-centric AI solution that generates multilabel organ segmentations to facilitate systemic TB whole-person research.The pipeline is based on nn-UNet and has the capability to segment 120 unique tissue classes from a whole-body 18F-FDG PET/CT image.
https://enhance.pet
GNU General Public License v3.0
196 stars 28 forks source link

FEAT: PET and CT Image Alignment with Affine Transformation #98

Closed LalithShiyam closed 5 months ago

LalithShiyam commented 8 months ago

Description

Requesting the reintroduction and enhancement of affine alignment for PET and CT images, a feature from MOOSE v1.0. This was requested here! This feature should align PET and CT images, output the aligned CT, and produce a multilabel segmentation aligned with the PET image. The primary use case is to facilitate the extraction of activity concentration from aligned multimodal images.

Justification

Accurate alignment and segmentation are critical for combining anatomical and functional imaging data, essential for precise activity concentration analysis in research and clinical applications.

Requirements

  1. Affine Alignment: Implement an improved affine alignment algorithm for PET and CT images (reuse from FALCON)
  2. Output Aligned Images: Provide the aligned CT image as output.
  3. Multilabel Segmentation: Output a multilabel segmentation map aligned with the PET image.
  4. Extract activitiesn: Use the multilabel segmentation map to extract activities from the PET image.

Requested by

Tagging @vikashg as they have requested the creation of this feature request to address their specific needs.

vikashg commented 8 months ago

Hi @LalithShiyam, Thanks for the feature request. So if I understand correctly Falconz is the tool that was used for registration and MOOSE v1.0 had a wrapper on it. BTW kudos to the team for coming up with acronyms :)

LalithShiyam commented 8 months ago

Hi @vikashg, Haha - I will convey it to the team! Exactly, we will be using FALCON :D! Also, leave us a star - if you find our work useful. Cheers, Lalith

LalithShiyam commented 5 months ago

Included in moosev3. 🎉