News

Keep up with the latest news from the NCI Center for Biomedical Informatics and Information Technology (CBIIT) and the data science communities.

A few of NCI’s Division of Cancer Biology grantees recently released publications on topics such as machine learning and artificial intelligence. These research results hold clues to how we research and develop various cancer treatments.

The new terminology for female reproductive neoplasms in NCI’s Thesaurus aligns with the latest World Health Organization standards. Other updates support CDISC’s standards and are intended to further facilitate the collection, management, and analysis of research data from human clinical trials and other studies.

Staff from CBIIT and NCI, alongside partners from NIH, FDA, and a consortium of scientists from across the world, joined forces to create reference samples and data call sets to help the cancer community further decipher cancer-related gene mutations. Their findings were recently published in Nature Biotechnology.

The NCI Cancer Research Data Commons’ Imaging Data Commons (IDC) has updated to include more features and 16 terabytes of medical imaging data files for cancer researchers and imaging informaticists.

The NCI Surveillance Research Program is hosting several webinars in October and November on resources for the analysis of SEER data and other cancer surveillance data. These webinars will include information on SEER*Stat, statistics, variables, methods, and software.

Dr. Jill Barnholtz-Sloan, CBIIT’s associate director of Informatics and Data Science and DCEG senior investigator, together with research colleagues, used a direct data matching approach to compare brain tumors in U.S. Veteran and non-Veteran populations. The study indicates that direct and deterministic data matching approaches have the potential to compare the distribution of tumors, treatment trajectories, and clinical outcomes of other cancers and rare diseases among these populations.

A new publication using NCI funding and resources shows that a machine learning model, called Panoptes, allowed cancer researchers to reliably predict subtypes of endometrial cancer. Such “computational pathology” offers a useful framework for supporting human pathologists, trimming the labor needed to interpret histological findings to under 4 minutes per slide, and eliminating the time and cost of genetic sequencing.

Interested in making data discoverable to the larger research community? Share your perspective with NIH, who wants to know how to improve data searchability and discovery. Cancer researchers, data submitters/generators, data users, and technology providers should respond by December 3, 2021.

The NCI Genomic Data Commons now has two new projects from studies about the potential health effects of exposure to ionizing radiation from the 1986 accident at the Chernobyl nuclear power plant in northern Ukraine.

Cancer researchers and data scientists have the opportunity to provide NIH and the FDA input on the requirements for accelerating clinical applications of next generation sequencing and radiomics (including those using artificial intelligence and machine learning). Responses to these Requests for Information are due Monday, November 1, 2021.