SU‐E‐T‐80: A Study of the Effect of Dose Grid Resolution On Stereotactic Plan Evaluation as a Function of Target Size

Abstract

Purpose: To investigate and quantify the impact of dose‐calculation grid‐size on dose volume histogram (DVH) for small targets, such as those in stereotactic radiosurgery (SRS). Methods: A cohort of ten patients with multiple metastasis was used in this planning study. All patients were treated with SRS and had target volumes ranging from 0.1 to 4.3cc. The Varian eclipse planning system was used for planning using a mono‐isocentric VMAT technique with four arcs. Each plan was calculated twice using a 1×1×1 mm3 and 3×3×3 mm3 dose grid respectively. Moreover, we used an open source DICOM‐RT viewer to read the raw‐RDose data (1×1×1 mm3 resolution) and were subsequently rescaled by linear interpolation to dose grids of 1.5×1.5×1.5, 2×2×2, 2.5×2.5×2.5 and 3×3×3 mm3. DVHs for all targets were recalculated and inter‐compared for the native and post‐processed resolutions to quantify the effect on DVH of dose grid resolution and correlate that with the PTV volume size. Results: DVHs shape and numerical values have a significant dependence on dose grid size resulting to a reduced Dmean, Dmax and Dmin with volume averaging effect being the main reason for these changes. Moreover, the magnitude of the DVH alteration is inversely proportional to tumor volume size. As the dose grid increases from 1×1×1 to 3×3×3 mm3, the DVH‐derived Dmean value is underestimated by ∼4, ∼7 and ∼10 % for tumors with volume sizes of 1.3, 0.5 and 0.1 cc respectively. Conclusion: The dose grid size that is used for the final dose calculation is affecting significantly the DVHs of small PTVs. The higher the dose grid size, the more noticeable is the DVH alteration, especially for tumors with a volume less than 1cc. For SRS planning, a dose grid of 1×1×1 mm3 should be used, especially for tumors with a volume < 1cc, to avoid erroneous DVHs. UTHSCSA, division of medical physics, has a research grant with Brainlab