MR-linac daily semi-automated end-to-end quality control verification

Area of Science:

• Medical Physics
• Radiation Oncology
• Quality Assurance

Background:

• Adaptive radiation therapy (ART) on integrated magnetic resonance (MR)-linacs requires rigorous quality assurance (QA) for accurate treatment delivery.
• Verifying data transfer and the performance of critical machine subcomponents is essential for the ART workflow.

Purpose of the Study:

• To develop and validate a novel automated end-to-end test suite for ART on MR-linacs.
• To verify data transfer across software platforms and quantify the performance of machine subcomponents.

Main Methods:

• A software tool was developed to quantify MR/megavoltage (MV) isocenter coincidence, couch positioning, isocenter shift accuracy, and Multi-Leaf Collimator (MLC)/jaw position accuracy.
• The test utilized a reference plan with simulated shifts on a phantom with MR/MV fiducials, executed within the adapt-to-position (ATP) workflow.
• Descriptive statistics and sensitivity measurements were used to quantify uncertainties and evaluate algorithm performance.

Main Results:

• The test suite demonstrated consistent MR-to-MV alignment tracking, detecting changes as small as 0.2 mm.
• Couch position consistency was within [-0.31, 0.26 mm].
• Accuracy for phantom shifts was within ±1.5 mm, jaw position differences averaged 0.27 mm, and MLC positions averaged -0.17 mm. Introduced errors were detected within 0.55 mm.

Conclusions:

• The developed end-to-end test suite is accurate and sensitive for detecting errors in ART workflows, including data transfer and machine subcomponents.
• The automated test captures system communication issues before patient treatment.
• The test's reliance on a standard phantom allows for potential expansion to other MR-linac sites.