Encrypted login | home

Program Information

A Multileaf Collimator and Jaw Symmetry Quality Assurance Technique Using An Electronic Portal Imaging Device

S Lebron

S Lebron*, B Lu , G Yan , D Kahler , C Liu , University of Florida, Gainesville, FL


SU-I-GPD-T-453 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose: Develop an efficient quality assurance (QA) technique that verifies multileaf collimator (MLC) relative leaf offsets and jaw/MLC symmetry using an electronic portal imaging device (EPID).

Methods: Two asymmetric fields with 0ᵒ (10cm MLC field width and X1=20cm, X2=3cm jaws) and 180ᵒ (10cm MLC field width and X2=20cm, X1=3cm jaws) collimator angles were utilized and analyzed individually. Three leaves adjacent to central axis were moved in by 1cm symmetrically for isocenter coordinate determination and jaw symmetry verification. A profile approximately 5mm from the crossline field edge was scanned across the three leaves in the inline direction. This was analyzed to determine the absolute location of the isocenter inline coordinate. The crossline profile across isocenter was used to determine the isocenter crossline coordinate. The isocenter’s crossline coordinate on both images should coincide; otherwise the asymmetric offset will be determined. Once isocenter was located, X1-X2 width was determined by calculating the edges of the inline profile passing through isocenter. Crossline profiles were scanned along the center of each leaf. The profiles’ edges determine the location of the leaves’ edges. Leaf offsets were calculated by subtracting each leaf position from the position of the leaf adjacent to isocenter. All profile edges were calculated by first parameterizing the profile’s penumbra region with a sigmoid equation. The technique was tested against two established methods (EPID and diode-array based), with different MLC systems, field widths and modalities (flattened and unflattened).

Results: Differences between the proposed and established methods were <0.5mm for two different collimator systems. Leaf offset deviations for different MLC openings were <0.3mm. Differences between flattened and unflattened beam offsets were <0.1mm. For pattern fields, differences between predetermined and calculated offsets were <0.5mm.

Conclusion: The proposed method is an efficient and accurate method of calculating relative leaf offsets and jaw/MLC symmetry for different MLC systems.

Contact Email: