A new artificial intelligence (AI) tool developed by a team at the University of Toronto may be able to significantly reduce the time needed to develop radiation therapy treatment plans for people with cancer.
The research published in the journal Medical Physics used AI to mine historical radiation therapy data and designed algorithms to develop recommended treatment strategies.
To check the AI-produced relevant treatment plans, the researchers looked at 217 patients with head and neck cancer who had their radiation therapy schedules developed via conventional methods. The plans were comparable.
The purpose of this study was to automatically generate radiation therapy plans for oropharynx patients by combining knowledge?based planning (KBP) predictions with an inverse optimization (IO) pipeline.
Radiation therapy treatment plans
They developed two KBP approaches the bagging query (BQ) method and the generalized principal component analysis?based (gPCA) method, to predict achievable dose–volume histograms (DVHs).
These approaches generalize existing methods by predicting physically feasible organ?at?risk (OAR) and target DVHs in sites with multiple targets. Using leave?one?out cross-validation, we applied both models to a large dataset of 217 oropharynx patients.
The predicted DVHs were input into an IO pipeline that generated treatment plans (BQ and gPCA plans) via an intermediate step that estimated objective function weights for an inverse planning model.
The KBP predictions were compared to the clinical DVHs for benchmarking. To assess the complete pipeline, we compared the BQ and gPCA plans to both the predictions and clinical plans.
To isolate the effect of the KBP predictions, we put clinical DVHs through the IO pipeline to produce clinical inverse optimized (CIO) plans. This approach also allowed us to estimate the complexity of the clinical plans.
The BQ and gPCA plans were benchmarked against the CIO plans using DVH differences and clinical planning criteria. Iso?complexity plans (relative to CIO) were also generated and evaluated.
"There have been other AI optimization engines that have been developed, but the idea behind ours is that it more closely mimics the current clinical best practice," said Aaron Babier, the lead author of the research from the University of Toronto Engineering Department.
At the moment, developing radiation therapy plans for each patient's tumor can take days, valuable time for patients as cancer often continues to grow and evolve, but also for physicians spending time designing these complex treatment strategies.
Head and neck cancers are notoriously difficult to design treatment plans for as tumors can be remarkably different from patient to patient. The researchers hope that as the tool worked so well on this tricky, complex cancer type, it should be able to handle more common tumor types that do not exhibit quite as much variation, such as prostate cancer.