Click here to


Are you sure ?

Yes, do it No, cancel

Development of a Gantry-Resolved EPID-Based Frame-By-Frame Patient-Specific VMAT QA Method

P Cardoso1*, Q Wu2 , B Liang3 , R Wei3 , (1) Duke Kunshan University, Kunshan, China, (2) Duke University Medical Center, Durham, NC, (3) Image Processing Center, Beihang University, Beijing, China


(Tuesday, 7/16/2019) 1:45 PM - 3:45 PM

Room: Stars at Night Ballroom 4

Purpose: Develop a gantry-resolved EPID-based method for patient-specific QA using integrated acquisition in TrueBeam clinical mode, and test it on large VMAT fields.

Methods: We developed a QA method using MATLAB that creates gantry-resolved pseudo-3D dose distributions from stacked portal images acquired on TrueBeam Linacs using dosimetry (integrated) acquisition mode. Predicted portal dose distributions were calculated based on MU information contained in XIM image headers. An independent calibration procedure was established, including a 2D profile correction map. Gamma index calculations were performed with a two-step calculation procedure that resamples steep dose gradient regions for improved calculation accuracy. Post-processing was used to address two acquisition artifacts: header ΔMU variations that do not reflect signal intensity variations, and banding patterns on acquired frames. Gantry angle accuracy requirement is given by angle-to-agreement (ATA) criteria. The method’s performance was assessed with gamma analysis (3%/2mm/3˚/10%threshold criteria) of 35 large VMAT fields using two EPID models and two photon energies.

Results: The improved gamma index calculation increased pass rates by an average of 2.8%, with many cases displaying increases of 6% to 8%. MU variations were reduced to less than 1% for irradiations with a constant dose rate, and smoothing effects on variations caused by dose rate changes were small (up to 1.5%). Banding patterns on central portions of the fields were almost eliminated without smoothing field borders. The resulting average pass rates were 95.4%, 97.8%, 98.6% and 96.2%, and no result was inferior to 90%. Analysis took around 6 to 8 minutes per field.

Conclusion: Our method presented excellent results for gantry-resolved gamma index analysis of VMAT fields. The two-step improved gamma calculation greatly improved pass rates and reduced variation in the results. Processing of acquisition artifacts enabled our method to be used in clinical dosimetry acquisition mode without the need of external acquisition hardware.


Quality Assurance, Electronic Portal Imaging, Temporal Resolution


TH- External beam- photons: Quality Assurance - VMAT

Contact Email