SBIR Phase II: A cloud-native, data-driven platform for automated quality assurance of radiation oncology treatment planning

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase II project is to improves patient outcomes in radiotherapy treatment. Healthcare providers are recognizing the growing imperative to lower costs and maintain quality of care by harnessing the digital health records available. In radiation oncology, only a small fraction of the valuable data produced during routine patient care is leveraged for the care of new patients. High-dose curative treatments frequently lead to side effects, such as dry mouth in head and neck cancer treatment, and rectal bleeding in prostate treatments. An optimized treatment plan could minimize complications. The proposed project will advance an information system to optimize and personalize radiation treatments. The proposed project leverages radiotherapy data to optimize treatment plans. Dose to organs at risk (OARs) must be kept low while ensuring the prescribed dosimetric coverage of the targeted disease. Intensity modulated radiotherapy (IMRT) treatments require significant optimization, creating extensive planning processes. This project will advance a treatment plan database for a data-driven clinical decision support solution that can predict achievable dose levels to each OAR. These levels can be used as intelligent optimization objectives by the treatment planning system (TPS) and as plan evaluation criteria. The system will be built for continuous refinement through ongoing augmentation with new patient plans, thus improving prediction performance with ongoing use, and allowing each clinical site to base predictions on its own data. In addition, integrated peer review tools enable physicians to efficiently share information and get additional clinical insight. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.