Program Information
Establishing a Risk-Based Tomotherapy MLC QA Program
Q Chen*, University of Virginia, Charlottesville, VA
Presentations
MO-DE-FS1-5 (Monday, July 31, 2017) 1:45 PM - 3:45 PM Room: Four Seasons 1
Purpose: To establish a risk-based tomotherapy MLC QA program that addresses the MLC fail modes based on its impacts the delivered dose accuracy.
Methods: Patients treated on two tomotherapy machine over the past year (728 cases in total) were analyzed. The actual leaf open time (LOT) was computed from the exit detector RawData during patient specific QA delivery. Error characteristics, including the dependency with gantry period (GP), planned LOT, systematic and random components were analyzed. A sinogram filled with uniform random numbers is setup as the routine MLC QA procedure to measure and monitor the MLC performance. The MLC performance characteristics are incorporated into an in-house 2nd dose calculation engine to predict the dose delivered (A process called “Virtual IMRT QA”). The predicted dose is compared with ion chamber measured dose.
Results: The systematic LOT error is found to correlate with planned GP, especially at the shortest GP (12 seconds) where the average LOT error reached to 1-2.5%. The error is also worse near either the shortest or the longest LOT. This agrees with our knowledge that the tomotherapy binary MLC has difficulty close or open for very short amount of time (~30 milliseconds). Through dosimetry analysis, the random component of the error has little effect to the dose. Therefore, the systematic component is the main focus of our QA program. We found that the worst dose error caused by MLC leaf closing is 5%. The virtual IMRT QA can catch this error with similar accuracy as the measurement.
Conclusion: We developed a routine MLC QA as well as a “Virtual IMRT QA” tool to monitor and manage the tomotherapy MLC performance issue. We believe this is an important step toward an efficient and effective quality management program for Tomotherapy.
Contact Email: