SU‐E‐J‐66: Improvement in SRS Commission and Patient‐Specific Quality Assurance

Abstract

Purpose: A linac‐based SRS system with a stereotactic mask was implemented. This work aims at selecting a proper detector for small‐field beam commissions, setting up a specially designed phantom for patient‐specific plan QA, and validating the dosimetric accuracy of SRS plans based on the Varian MLC‐120 and BrainLAB cones.Methods: Four different detectors — diode, micro ion chamber, IBA chamber, and Farmer chamber, were evaluated for small field commissioning. PDD, profile, and output measurements of fields from 5×5mm to 150× 150mm were compared and discussed. A micro‐chamber was selected based on its good spatial resolution (0.016cc sensitive volume) and higher stability and sensitivity (3 and 1.5 times greater, respectively, than a diode). To reduce measurement uncertainties, both PDD and profile scans were iterated (more iterations were used for smaller fields) and summed up using commercial software. The shift of reference‐field readings was within 0.3% during the output measurements. A Lucy phantom was set up in the SRS frame and CT scanned. Treatment plans were mapped to the CT image and measured with a calibrated micro‐chamber. The implemented SRS system was validated by comparing the planned and measured doses for targets with different volumes and shapes.Results: For a typical SRS target (1.26cc), the dose differences were 0.71% for a dynamic arc/MLC plan and 0.75% for an arc plan using a 20mm cone. For a small and spherical target (0.50cc), the differences were 2.13% for a dynamic arc/MLC plan and 0.17% for an arc plan using a 15mm cone. For a very small and irregular target (0.35cc), the differences were 2.60% for a dynamic arc/MLC plan and 4.72% for an arc plan using a 10mm cone Conclusions: Considering its dosimetric accuracy for different targets and the availability of a safety‐interlock, MLC based SRS treatment is highly recommended.