Purpose: To develop an open-source software application that allows for the storage, search, and downloading of DICOM images to support data sharing, enabling imaging informatics applications and data mining for cancer research.
Methods: The system utilizes a Java server that provides REST services to independent applications such as the Angular-based graphical user interfaces. The system also includes a Solr server that indexes all tags in the DICOM headers and allows rapid retrieval based on any of the information in the tags. The applications are loosely coupled, enabling independent upgrades, speeding the delivery of features to users.
Results: The National Biomedical Imaging Archive (NBIA) provides web-based and programmatic access to DICOM images by using role-based security. NBIA consists of a search client, data administration tools, user administration tools, and the NBIA Data Retriever, which supports downloading of images to the user’s local computer. NBIA enables several medical imaging archives, including The Cancer Imaging Archive (TCIA) of the National Cancer Institute (NCI), which currently hosts over 27.5 million images and empowers downloads of over 100 TB per month.NBIA incorporates the Clinical Trial Processor software from RSNA and supports customized de-identification of images. NBIA can integrate with other applications through a set of APIs that can facilitate data query and data transfer between multiple repositories, such as between the NCI image repositories, TCIA, and Imaging Data Commons, a cloud-based resource. The NBIA code is available on GitHub. Technical documentation including an API guide can be found on the NCI’s NBIA wiki page.
Conclusion: NBIA is an open-source software application that enables the creation of secure and searchable archives of medical images in DICOM format. NBIA is a constantly evolving system that takes advantage of new technologies and approaches to meet the needs of the cancer imaging community.