New results from a study partially funded by NCI could help you and other clinicians improve cancer treatment planning processes and concentrate on more complex cases. The study shows the power of machine learning and the promise of fully automated radiation therapy planning when researchers created and validated many KBP models across ten cancer treatment fields. This research highlights that KBP models can generate consistent, high-quality plans, decreasing planning time and reliance on individual expertise.

The models underwent testing on multiple cancer locations, such as prostate, bladder, breast, and lung. Results suggest that automated planning could standardize procedures and reduce human mistakes. Researchers designed these models to formulate radiation treatment plans, eliminating the need for manual input from treatment planners.

Physicians found that 88% of the generated plans were “acceptable as is,” while 98% were acceptable after slight adjustments. These results indicate that automation could significantly lighten the workload for cancer treatment planners and improve overall efficiency in radiation therapy.