Table 2 Summary of the AI4HI projects, listing their goals, use-cases, types of metadata identified so far
From: Position of the AI for Health Imaging (AI4HI) network on metadata models for imaging biobanks
Project | Goal | Considered use cases | Types of metadata | Adopted models |
|---|---|---|---|---|
PRIMAGE | To build an imaging biobank for the training and validation of machine learning and multiscale simulation algorithms | Paediatric neuroblastoma and diffuse intrinsic pontine glioma | DICOM tags Image analysis metadata (registration, denoising, radiomics) Clinical variables | DICOM-MIABIS OMOP CDM |
EuCanImage | To build a European cancer imaging platform for enhanced AI in oncology | Eight use cases regarding liver, breast, and colorectal cancer | Imaging data Clinical variables | DICOM-MIABIS ICGC-ARGO |
INCISIVE | To improve cancer diagnosis and prediction with AI and big data | Lung, breast, colorectal, and prostate cancer | Imaging data Clinical and biological data | FHIR |
CHAIMELEON | To develop a structured repository of health images and related clinical and molecular data | Lung, breast, prostate, and colorectal cancer | Imaging data Clinical variables | DICOM-MIABIS OMOP CDM |
ProCancer-I | To develop an AI Platform integrating imaging data and models | Prostate cancer | Imaging data Clinical variables | DICOM-Radiation therapy OMOP CDM with Oncology Extension |