Influence of Calculation Grid Resolution on Monte Carlo Dose Estimation In Cranial Radiotherapy
Abstract
Purpose
To assess the effect of Monte Carlo (MC) grid resolution on dose estimation in cranial radiotherapy across different tumor types.
Methods
110 cranial treatment plans generated with Elements Cranial were evaluated, including 10 arteriovenous malformations (AVM), 20 glioblastomas (GBM), 20 meningiomas (MGM), 20 brain metastases (BM), 20 pituitary adenomas (PA), and 20 vestibular schwannomas (VS). Analyzed parameters included GTV and PTV volumes and dimensions, PTV margins, and the degree of PTV–bone overlap. Dosimetric endpoints consisted of D98%, D95%, D2%, and mean dose. MC calculations were performed using grid resolutions of 1%/1mm and 2%/2mm.
Results
GBM cases exhibited the largest GTV and PTV volumes, whereas VS cases presented the smallest target volumes and the highest proportion of PTV overlapping bone structures. No statistically significant differences in dose metrics were observed between grid resolutions for AVM, GBM, or MGM. In contrast, the use of a 2%/2mm MC grid resulted in significant dose reductions for BM, PA, and VS across all evaluated coverage metrics. Specifically, significant differences were observed for BM (D98%, D95%, and D2%; p=0.002), PA (D98% and D95%; p=0.002; D2%; p=0.027), and VS (D98%; p=0.019; D95%; p=0.013; D2%; p=0.002). After adjustment for PTV volume and PTV–bone intersection, median dose values remained lower with the 2%/2mm grid compared to the 1%/1mm grid for D98% (25.27Gy vs. 25.60Gy; adjusted p=0.026), D95% (26.14Gy vs. 27.05Gy; adjusted p=0.005), and D2% (30.00Gy vs. 30.48Gy; adjusted p<0.001). Additionally, higher conformity and gradient index values were associated with the coarser grid, particularly in smaller or geometrically complex targets.
Conclusion
MC grid resolution significantly affects dose calculation in small or anatomically irregular cranial targets, including BM, PA, and VS. The 2%/2mm grid produces systematic dose reductions and increases in conformity and gradient indices, underscoring the importance of grid selection in high-precision cranial treatment planning.