Program Information
MLC Consistency and Patient Specific VMAT QA with Log Files
R McDermott*, A Godley , S Balik , The Cleveland Clinic, Cleveland, OH
Presentations
SU-I-GPD-T-476 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: To test the consistency of MLC positions throughout the day, to study the affect treatment complexity has on Gamma passing rates and determine standards for log file based patient specific VMAT QA.
Methods: For 20 VMAT patients we compared the Gamma between measured and planned treatments using two methods. The phantom method uses a 2D array of ion chambers. The measured dose is compared to the planned dose using the 3%/3 mm Gamma criterion with a minimum passing rate of 90%. The log file method compares the planned fluence map provided to the delivered fluence map calculated from the log file by RadCalc. The passing rate was compared between the two methods. Log files were also collected for each treatment delivery. We looked for an increase in error in MLC positions as a function of time of day. MLCs were initialized each morning. With 30-40 VMAT patients treated daily, the accumulated MLC position error could affect treatment delivery. Lastly we analyzed the relationship between Gamma passing rates and plan complexity, where MU was used as surrogate for complexity.
Results: There was no correlation between MLC error and time of day, R² of 0.0164 for 95th percentile error and 0.019 for max RMS error. There is a negative correlation between Gamma passing rate and plan complexity, R² of 0.27. To give passing rates consistent with phantom measurements, which suffer from setup errors and low resolution, log file based QA should use 2%/2 mm at 90%.
Conclusion: With daily initialization of MLC, there was no effect of MLC position errors. As the complexity of a plan increases, the Gamma passing rate decreases. The log file method is valid to QA the MLC positions for patient specific VMAT treatments. However it relies on the accuracy of the log files.
Contact Email: