NRG Oncology Network Group Operations Center

Nrg Oncology Foundation, Inc., Philadelphia PA

Investigators

Linked publications, trials & patents

Abstract

NRG Oncology Operational Ontology for Oncology (O3) towards AI/ML Ready Standardization Charles Mayo, PhD, Supplement Principal Investigator Project Summary The parent grant, NRG Oncology, was formed as a collaborative effort between the former National Surgical Adjuvant Breast and Bowel Project (NSABP), the Radiation Therapy Oncology Group (RTOG), and the Gynecologic Oncology Group (GOG) with the goal of improving the lives of cancer patients through practice- changing multi-institutional clinical and translational research, with emphases on gender-specific malignancies and localized or locally advanced cancers of all types. This proposal outlines a supplement project that aims to develop user-friendly tools to enable the seamless conversion between different data standards for radiotherapy data elements collected during NCI clinical trials. The project targets the standardization and interoperability of radiotherapy data, which is crucial for facilitating the use of machine learning and artificial intelligence in the field. The project's specific aims include developing a comprehensive mapping and conversion framework between OOO/FHIR, CDISC, OMOP, and NCI Thesaurus standards, designing and implementing user-friendly tools for data aggregation and seamless conversion, and pilot testing and validating the developed tools with collaborating institutions and NCI clinical trial data. The team will coordinate with leadership from these standards groups to ensure accuracy in mapping and concept definitions. The tools developed will incorporate data validation and error-handling mechanisms to ensure data quality and integrity during the conversion process. Ultimately, the project aims to promote the adoption of these tools by the broader radiation oncology community, specifically for NCI clinical trials, and contribute to improving cancer patient outcomes and advancing cancer research. The aims of the project are: Aim 1 - We will extend our prior work with O3, providing easily accessible, publicly available mappings between O3, OMOP, CDISC, NCI Thesaurus and HL7-FHIR supporting clinical trials for cancer care. Aim 2 – We will extend our tools for data extraction supporting mapping and connection of O3 based Artificial Intelligence (AI) ready data sets to machine learning algorithms. Aim 3 The tools developed in aim 2 will be used by a set of institutions to demonstrate the ability to aggregate RTOG 0617 information for a representative (5 patients/institution) set of patients. Aim 4 Mappings and tools will be published on a publicly accessible GitHub repository.