Program Information
Modulation Index for VMAT
J Park1*, S Park1 , J Kim1 , H Kim2 , J Kim1 , J Carlson3 , S Ye3 , (1) Seoul National University Hospital, Seoul, (2) Yonsei University, Seoul, (3) Seoul National University, Seoul,
Presentations
SU-E-T-550 (Sunday, July 12, 2015) 3:00 PM - 6:00 PM Room: Exhibit Hall
Purpose: To present modulation indices (MIs) for volumetric modulated arc therapy (VMAT).
Methods: A total of 40 VMAT plans were retrospectively selected. To investigate the delivery accuracy of each VMAT plan, gamma passing rates, differences in modulating parameters between plans and log files, and differences between the original plans and the plans reconstructed with the log files were acquired. A modulation index (MIt) was designed by multiplications of the weighted quantifications of MLC speeds, MLC accelerations, gantry accelerations and dose-rate variations. Textural features including angular second moment, inverse difference moment, contrast, variance, correlation and entropy were calculated from the fluences of each VMAT plan. To test the performance of suggested MIs, Spearman's rank correlation coefficients (r) with the plan delivery accuracy were calculated. Conventional modulation indices for VMAT including the modulation complexity score for VMAT (MCSv), leaf travel modulation complexity score (LTMCS) and MI by Li&Xing were calculated, and their correlations were also analyzed in the same way.
Results: The r values of contrast (particular displacement distance, d = 1), variance (d = 1), MIt, MCSv, LTMCS and MI by Li&Xing to the local gamma passing rates with 2%/2 mm were 0.547 (p < 0.001), 0.519 (p < 0.001), -0.658 (p < 0.001), 0.186 (p = 0.251), 0.312 (p = 0.05) and -0.455 (p = 0.003), respectively. The r values of those to the MLC errors were -0.863, -0.828, 0.917, -0.635, -0.857 and 0.795, respectively (p < 0.001). For dose-volumetric parameters, MIt showed higher statistically significant correlations than did the conventional modulation indices.
Conclusion: The MIt, contrast (d = 1) and variance (d = 1) showed good performance to predict the VMAT delivery accuracy showing higher correlations to the results of various types of verification methods for VMAT.
Funding Support, Disclosures, and Conflict of Interest: This work was in part supported by the National Research Foundation of Korea (NRF) grant (no. 490-20140029 and no. 490-20130047) funded by the Korea government.
Contact Email: