The Emory Knee Radiograph (MRKR) dataset is a large, demographically diverse collection of 503,261 knee radiographs from 83,011 patients, 40% of which are African American. This dataset provides imaging data in DICOM format along with detailed clinical information, including patient-reported pain scores, diagnostic codes, and procedural codes, which are not commonly available in similar datasets. The MRKR dataset also features imaging metadata such as image laterality, view type, and presence of arthroplasty, enhancing its value for research and model development. Data are de-identified and available publicly and openly under a CC-BY-SA license. This dataset provides an opportunity for researchers studying osteoarthritis and pain to develop equitable deep learning models to help improve patient outcomes and pain.
Price et al. (2026) studied this question.