Loading
collectionsCollection
The Cancer Imaging Archive (TCIA)
The public archive of cancer scans that most radiology AI is trained and tested on.
Overview
The Cancer Imaging Archive (TCIA) holds de-identified cancer imaging, CT, MRI, PET and pathology, linked to TCGA and to clinical trials, and it is the main training and testing source for radiology AI models. It has over 200 collections, pairs with TCGA genomics for radiogenomics, and is accessed through the NBIA portal and the Imaging Data Commons. TCIA is run by the NCI with the University of Arkansas; most collections are CC BY, with terms set per collection. On OnCo it is tagged data, linked to the technology AI in radiology and the institution National Cancer Institute (NIH), and cited by the idea One certified open-source de-identification pipeline for scans and slides.
Holds
De-identified cancer imaging (CT, MRI, PET, pathology) linked to TCGA and clinical trials; the main training source for radiology models.
Licence
Mostly CC BY; per-collection
Maintainer
NCI / University of Arkansas
Dataset registry
Size: More than 200 imaging collections (CT, MRI, PET, pathology)
Access: open
Consent and reuse: De-identified; mostly CC BY, some collections restricted.
All datasets and the models trained on them →Similar pages
- CollectionNCI Imaging Data Commons (IDC)
Shares National Cancer Institute (NIH) and the tag data.
- CollectionNCI CDAS: NLST and PLCO screening trial data
Shares National Cancer Institute (NIH) and the tag data.
- TechnologyRadFM (generalist radiology foundation model)
Shares Radiology imaging as a data modality (CT, MRI, TCIA), AI in radiology.
- TechnologyCT-FM (whole-body CT foundation model)
Shares Radiology imaging as a data modality (CT, MRI, TCIA), AI in radiology.
- PersonMark Schiffman
Open-source projects
All open source in oncology →Open-source projects that are the code behind this collection or publish it, from OnCo's own catalogue: licence and last activity as the repository reported them on the day of the fetch. Listing is not endorsement; check the licence before reuse and the validation before clinical use.
Jupyter notebooks from The Cancer Imaging Archive team on finding, downloading and analysing its collections.
136 starslast push 2026-09-10The Cancer Imaging Archive, Frederick National LaboratoryRepositoryPaperA Python client for The Cancer Imaging Archive's NBIA API from the Haibe-Kains lab.