Room: AAPM ePoster Library
Purpose: The clinical workflow of MR-linac based adaptive radiotherapy requires prompt quality assurance of the online adaptive treatment plan (P?). We developed a computerized independent program (called CSITP), that checks the sanity of P?, and ensures the integrity of data transfer for a Unity MR-Linac system.
Methods: To replace the time-consuming dosimetric QA of P?, we use CSITP to calculate the similarity comparing P? to the original reference plan at simulation (P?) on a wide spectrum of dosimetric driven parameters, including beam angle, number of segments, segment area, MU weighted segmentation area, center of mass of the fluence, and beam/plan MU. Similarity between P? and P? was investigated for two clinical scenarios: “adapt to position” (rigid translational setup errors) and “adapt to shape” (translational + rotational setup errors and deformation). Setup shifts ranging 5-20mm in superior-inferior and anterior-posterior directions were simulated on prostate and liver plans to derive reasonable thresholds which trigger alerts. Approved P? is subsequently pushed from the planning system (Monaco) to the delivery and management system (Mosaiq), which hosts a separate database. CSITP automatically queries the two databases, and verifies that plans in the two systems are identical. Log files recording the checking activities are generated for future chart round review and analysis.
Results: The threshold for similarity between P? and P? is set to 5%/4mm and 15%/1cm for the clinical scenario of “adapt to position” and “adapt to shape”, respectively. CSITP can catch intentional setup and corruption errors introduced in testing. The execution of CSITP only takes a few seconds, imposing no burden on the entire treatment.
Conclusion: CSITP can perform independent checks of the sanity and integrity of the adaptive plan, and provide raw data for workflow and quality assurance analysis. Proper alerting criteria need further investigations with more accumulated patient data.