Program Information
Anatomically-Based Log File QA Using Monte Carlo as An Alternative Patient-Specific QA Technique
C Stanhope1*, J Liang2 , D Drake3 , D Yan4 , (1) William Beaumont Hospital, Royal Oak, MI & Wayne State University, Royal Oak, MI , (2) William Beaumont Hospital, Royal Oak, MI, (3) William Beaumont Hospital, Royal Oak, MI, (4) William Beaumont Hospital, Royal Oak, MI
Presentations
SU-I-GPD-T-284 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: To investigate the feasibility of a Log File (LF)/Monte Carlo(MC)-based QA system.
Methods: 23 clinical VMAT cases [46 arcs] previously planned in Pinnacle, calculated using Convolution-Superposition (CS), treated using an Elekta Agility MLC, and QA’ed using ArcCHECK were selected. Log files were converted into DCMRT plan format with one control point per useful log file sample. Elekta’s Log File Convertor R3.2 records linac delivery parameters (dose rate, gantry angle, leaf position) every 40ms. Dose was calculated on the patient’s anatomy for the original plan and log file plan using ScientificRT’s SciMoCa MC algorithm. Doses were labeled Plan-MC and LF-MC respectively. Three dose comparisons were made: (a) LF-MC vs Plan-MC isolates delivery error; (b) Plan-MC vs Plan-CS isolates calculation algorithm; (c) LF-MC vs Plan-CS combines the first two. 2%/2mm per-beam gamma pass rates were calculated for each comparison using 10% and 85% dose thresholds. Per-beam percent differences in OAR mean dose, PTV D₁, and PTV D₉₉ were calculated for each comparison.
Results: Plan-CS vs. ArcCHECK 2%/2mm pass rates were 97.2±2.2% (standard 10% threshold) and 91.1±18% (85% target threshold). Pass rates for comparisons a-c were 99.9±0.2%, 92.5±2.9%, and 92.3±3.0% for the 10% threshold. Using an 85% threshold, pass rates decreased to 99.8±0.1%, 80.0±11.7%, and 83.7±16.1%. Mean OAR doses were consistently within 1%, 3%, and 3%. Change in PTV D₁ was -0.2±0.3%, -0.1±1.3%, and -0.3±1.3%. Change in PTV D₉₉ was 0.1±0.4%, -4.2±2.9%, and -4.1±2.8%.
Conclusion: LF QA dose difference was largely due to calculation algorithm and minimally due to log file-determined delivery error. Increasing the gamma dose threshold yielded decreased passing rates for both QA techniques. However, whereas ArcCHECK consistently yielded passing (> 90%) results, LF QA recorded worrisome target coverage (gamma 85% and D₉₉).
Funding Support, Disclosures, and Conflict of Interest: This research is funded by an Elekta Research Grant
Contact Email: