Open access medical imaging datasets are needed for research, product development, and more for academia and industry.  We hope this guide will be helpful for machine learning and artificial intelligence startups, researchers, and anyone interested at all.  This is a growing list and will be periodically updated – if you know of another open medical imaging dataset, please email

Radiology Open Repositories:

  • NIH – 100,000 chest x-rays with diagnoses, labels, annotation
  • TCIA – The Cancer Imaging Archive consisting of extensive number of datasets from Lung IMage Database Consortium (LIDC), Reference Image Database to Evaluate Response (RIDER), Breast MR, Lung PET/CT, Neuro MRI scans, CT Colonoscopy, Osteoarthritis database (MIA), PET/CT phantom scans
  • OASIS – Cross sectional imaging MRI data
  • Kaggle Data Science Bowl 2017 – Lung cancer imaging datasets (low dose chest CT scan data) from 2017 data science competition
  • Stanford Artificial Intelligence in Medicine / Medical Imagenet – Open datasets from Stanford’s Medical Imagenet
  • MIMIC – Open dataset of radiology reports, based on critical care patients
  • National Library of Medicine MedPix – Free open source database of over 12,000+ cases
  • SMIR – Full Body CT Scans
  • SMIR – High Resolution Scapular Scan (CT)
  • SMIR – Temporal Bone CT scans
  • ADNI – Alzheimer’s Disease Neuroimaging Initiative with MR, PET images, genetics, cognitive tests, CSF and blood biomarkers
  • FITBIR – Federal Interagency Traumatic Brain Injury Research system with MRI, PET images of TBI
  • Cancer Digital Slide Archive – Pathology and Radiology images (excellent rad-path correlations) for GBM – Glioblastoma multiforme
  • Johns Hopkins Diffusion Tensor Imaging (DTI) / Lab of Brain Anatomi…– High resolution neuro-MRI scans
  • Grand Challenge – data from over 100+ medical imaging competitions in data science
  • MIDAS – Lupus, Brain, Prostate MRI datasets

In additional, image resources may span beyond actual datasets of X-Ray, MR, CT and common radiology modalities.

Non-Radiology Open Repositories (General medical images, historical images, stock images with open licenses):