National Cancer Institute (NCI) Imaging Data Commons: Towards Transparency, Reproducibility, and Scalability in Imaging Artificial Intelligence (AI)

The remarkable advances of Artificial Intelligence (AI) technology are revolutionizing established approaches to the acquisition, interpretation, and analysis of biomedical imaging data. Development, validation and continuous refinement of AI tools requires easy access to large, high quality, annotated datasets, which are both representative and diverse. The National Cancer Institute (NCI) Imaging Data Commons (IDC) hosts large and diverse publicly available cancer image data collections, with many of those collections accompanied by various forms of annotations, including volumetric segmentations, and clinical data. All of the data hosted by IDC is harmonized into uniform standard representation to achieve interoperability, ease reuse and data aggregation, and support improved transparency and data provenance. As a data commons, IDC places emphasis on enabling reuse, analysis and sharing of the analysis results by the community. To achieve this goal, IDC data is available within scalable cloud-based infrastructure that can simplify transparency and reproducibility of the analyses, while making it possible to achieve superior time- and cost-efficiency of the computation. In this talk I will discuss the current status of IDC and the potential of this publicly available resource to improve transparency, reproducibility and scalability in medical image computing research.