Skip to content

Med-ImageNet: A Standardized Resource for AI-Ready Oncology Imaging#

Core Features#

Med-ImageNet is an open-source platform that transforms heterogeneous cancer imaging collections into harmonized, AI-ready resources for oncology research. It provides tools to query, download, and preprocess medical imaging datasets from public and user-provided sources through a unified Python interface.

Index

Platform Components#

The platform comprises three integrated components:

  1. Med-ImageDB -- Dataset indexing, query API, and secure image and metadata retrieval across all supported collections. The index can be found here.

  2. Med-ImageTools -- Standardized preprocessing including DICOM ingestion, voxel harmonization, intensity normalization, and metadata alignment. The tools can be found here.

  3. Med-ImageNet Repository -- Unifies these modules into a scalable and reproducible data compendium supporting both raw data access and AI-ready outputs (e.g., NIfTI format) for deep learning integration.

Architecture

Installing Med-ImageNet#

pip install med-imagenet

imgnet --help

Key Capabilities#

  • Queries across all supported collections with associated metadata
  • Establishes explicit links between paired imaging modalities (e.g., CT with RTSTRUCTs)
  • Query and request datasets based on imaging region and imaging modality
  • Downloads from TCIA/IDC, S3, Dropbox, Zenodo, and HuggingFace sources
  • Processes raw DICOM files to generate AI-ready NIfTI outputs, tabular metadata files, and dataset summaries

License#

This project uses the following license: MIT License