This piece was originally published in the November 2016 issue of electroindustry.
David A. Clunie, MD, PixelMed Publishing
Cancer is the leading cause of death in many states. Its screening, diagnosis, treatment, and research relies on the interoperability of digital medical imaging systems. That fact is not lost on the National Cancer Institute (NCI), the federal government’s principal agency for cancer research and training, located within the National Institutes of Health (NIH).
One of the most visible NCI activities is The Cancer Image Archive (TCIA). TCIA consists of a publicly accessible repository of large collections of DICOM image studies of cancer patients. These collections are used by researchers to test new analysis techniques, engineers to develop new software, and academics to use as a teaching tool.
TCIA includes human, animal, and phantom images of many types of tumors. Like other agencies within the NIH, the NCI encourages open access and data sharing to expedite the translation of research results into knowledge, products, and procedures that improve human health. The use of DICOM allows commercial and open-source, off-the-shelf tools to load, display, and analyze medical images.
The NCI is also committed to supporting activities that improve the standards themselves. It has been a member of the DICOM Standards Committee since 1999. Over the last decade, the NCI has augmented three key areas of DICOM for cancer research:
Thorough de-identification (the removal of identifiable information from patients’ images) and other activities that involve sharing images beyond the purpose of their original acquisition are critical to the success of TCIA. With input from manufacturers, pharmaceutical companies, academics, researchers, and the NCI, DICOM developed a detailed description of de-identification requirements that have been integrated into the process of submission of images to the TCIA.
Encoding quantitative results
Cancer imaging entails assessment of disease extent and progression, traditionally a laborious, manual process. In order to develop imaging biomarkers for assessment of therapeutic efficacy, to relate the imaging phenotype to other data, and to improve automated analytic methods, it is necessary to share images and image-related quantitative results. This objective is demonstrated by the NCI’s Quantitative Imaging Network (QIN), a collaborative effort to promote the research and development of quantitative imaging methods, including DICOM. A related effort is that of Quantitative Imaging Informatics for Cancer Research (CIICR), which focuses on extending research software to implement DICOM support and extend DICOM to facilitate the translation of research techniques from the laboratory to the bedside.
Small animal identification
Cancer research goes beyond human imaging; it includes pre-clinical activities using small animals. Adoption of DICOM-dedicated devices has been slow and has acted as a barrier to data sharing and reuse. Additional metadata is required to document imaging and handling conditions that affects image interpretation. Accordingly, the Clinical and Translational Imaging Informatics Project (CTIIP) supported the establishment of the dedicated DICOM Working Group 30, which produced several extensions to the standard.
MITA is proud to be an active collaborator in the cancer informatics community by supporting efforts towards adoption and improvement of DICOM in clinical care and translational research.
DICOM (Digital Imaging and Communications in Medicine) is a global information technology standard that is used to ensure the interoperability of systems used to produce, store, display, process, send, retrieve, query, or print medical images.
MITA is the secretariat for the DICOM stanndard. NEMA holds the copyright. It is also known as NEMA PS3 and ISO 12052:2006.Read the November 2016 issue of electroindustry.