3135 - Impact of AAPM TG-51 Corrections on Dosimetric Leaf Gap and Multi-Leaf Collimator Modeling Parameters: A Multi-Energy Statistical Analysis

Purpose/Objective(s): The traditional dosimetric leaf gap (DLG) measurement for Linac commissioning uses raw ion chamber readings without mandatory application of AAPM TG-51 correction factors (P_ion, P_TP, P_elec, P_pol). This practice may introduce systematic errors and reduce cross-institutional comparability. This study quantifies, for the first time, the statistical impact of applying full TG-51 corrections to both DLG and the linear regression slope, a critical parameter for multi-leaf collimator (MLC) transmission modeling, across five clinically relevant photon energies using a rigorous, repeatable measurement framework.

Materials/Methods: Measurements were performed on an SRS-dedicated Linac using a water phantom and an ion chamber. For each energy (6 MV, 6 FFF, 10 MV, 10 FFF, 16 MV), the chamber was placed at isocenter (90 cm Source-to-Surface Distance (SSD) at 10 cm depth in a 10×10 cm² reference field. DLG was determined via the standard closed-MLC slit field method with gap widths of 2-20 mm. Four repeated measurements per energy were acquired to assess reproducibility. For each raw reading M_raw, the fully corrected dose M_corr was calculated as M_raw×P_ion×P_TP ×P_elec×P_pol. DLG (absolute Y-intercept) and slope were extracted from linear regression of corrected and uncorrected datasets. Paired two-tailed t-tests (a=0.05) were performed to evaluate statistical significance (corrected vs uncorrected) per energy and globally (n=20).

Results: DLG corrections yielded no statistically significant difference for 6 MV, 10 MV, 10 FFF, and 16 MV (p=0.39). However, 6 FFF showed a highly significant increase of +0.0187 mm (p<0.001), representing a 3.2% relative change. In contrast, the slope was significantly altered for every energy: 6 MV: +0.0372 (p<0.001, +0.47%), 6 FFF: +0.0225 (p<0.001, +0.27%), 10 MV: +0.0340 (p<0.001, +0.49%), 10 FFF: –0.0169 (p<0.001, –0.23%), and 16 MV: +0.0130 (p<0.001, +0.20%). Overall pooled analysis confirmed a highly significant global effect on slope (mean ?=+0.0180, p<0.001). The global effect sizes were consistent across repeated samples, confirming measurement stability.

Conclusion: TG-51 corrections do not uniformly affect raw DLG values; flattened beams and 10 FFF were insensitive, whereas 6 FFF DLG was systematically underestimated without correction, a clinically relevant bias for SRS, SBRT, and VMAT plans using small MLC gaps. Conversely, slope is universally and significantly impacted, with energy-dependent direction and magnitude. These findings establish that omission of TG-51 corrections during commissioning introduces energy-specific systematic errors in MLC modeling that propagate into treatment planning system (TPS) dose calculation accuracy. We strongly recommend application of full TG-51 dosimetric corrections to all ion chamber-based DLG and slope measurements and propose this multi-energy statistical framework as a new quality benchmark for modern Linac commissioning.