Room: Exhibit Hall | Forum 4
Purpose: Motivated by limitations of conventional gamma analysis, we evaluate a modified approach to patient specific VMAT QA. The sensitivity of a novel Enhanced Gamma Index (EGI) method to common treatment planning and delivery errors is evaluated.
Methods: Clinically approved VMAT plans for 20 patients (5 each: brain, head and neck, lung and prostate) were included in the study. Plans were re-optimized with changes to PTV/OAR dose constraints so as to render each plan clinically unacceptable (OAR Dmax, Dmean above threshold and/or sub-optimal PTV coverage). Original versus modified treatment plans were compared via conventional and proposed QA methods based on 95% points passing 5%/5mm, 3%/3mm and 1%/1mm (dose diff%/DTA) criteria. Results of gamma analysis were correlated with corresponding changes in PTV/OAR doses.
Results: For a HN patient, a 28% increase in spinal cord Dmax was not detectable with conventional QA (>97% pass rate @ 3%/3mm). However, with the structure-specific EGI QA, spinal cord failed gamma criteria (89% pass @ 3%/3mm and 44% pass 1%/1mm). Similarly, for a brain patient, a 5.8% increase in optic-nerve dose beyond threshold was not detected conventionally but was readily flagged by EGI QA (83% and 42% pass @ 3%/3mm & 1%/1mm respectively). A lung-patient plan with sub-optimal PTV coverage was similarly triggered during modified QA (78% pass @ 3%/3mm) but missed in conventional QA. Lastly, a prostate patient, with high rectal dose passing conventional QA was captured in the modified approach even with relaxed gamma criteria (28% pass @ 3%/3mm, 74% @ 5%/5mm)
Conclusion: The limitations of conventional gamma analysis based QA methods have been documented. The proposed QA approach that couples gamma test with spatial dose information was evaluated and found clinically useful. The method lends itself to evaluating clinically relevant parts of treatment plan for follow-up action.