Edit this page

NA-MIC Project Weeks

Back to Projects List

Investigate MONAI generative modeling for Imaging Data Commons

Key Investigators

Presenter location: In-person

Project Description

Generative learning refers to a class of techniques that process large amounts of training data into models that can be used for a variety of tasks such as synthetic data generation, image compression, enhancing resolution, classifying images, and content based retrieval. Recently a generative package has been added to the open source MONAI software.

This project will explore the application of MONAI generative tools to data on the NCI Imaging Data Commons.

Objective

  1. Study the existing material and collect information from other interested parties
  2. Make plans about what experiments would be interesting
  3. If possible do some small experiments to better understand what’s possible and what effort and resources would be required to scale up

Approach and Plan

  1. Explore creating an IDCDataset compatible with MONAI Datasets using idc-index to fetch data
  2. Investigate adapting tutorial code to work with IDC data
  3. Try running some small tests, such as running the superresolution tutorials on IDC data
  4. Document how IDC can be used with MONAI for research

Progress and Next Steps

  1. Discussed the project with people at project week for feedback
  2. Contacted Mark Graham of KCL, a MONAI generative researcher/developer for advice
  3. Implemented a first pass combination of IDC data with MONAI generative notebook
  4. Ran tests on colab and workstations
  5. Adapted example data (8-bit) to dicom (16-bit) data to accomodate dynamic range differences
  6. Explored parallel and federated approaches

Illustrations

No response

Background and References