Which Is an Optimal Dosimeter to New Optical CT Scanner (P3DS) Between BANGkit and PRESAGE? A Feasibility Study for Brain SRT Case

Abstract

Purpose/Objective(s): For adaptive radiation therapy (ART), delivered dose evaluation is important for deriving the replanning strategy during treatment. Planned dose shifting method provides a fast and simple way to estimate the daily delivered dose while its effectiveness for dose assessment is in doubt. This study aims to compare this method to the currently-in-use dose recalculation method in different dose evaluation approaches. Materials/Methods: Five head and neck (H&N) cancer patients who received tomotherapy with a total of 170 pretreatment megavoltage CTs (MVCTs) were included. For each patient, contours on each day MVCT were deformed from planning images by deformable image registration (DIR). Daily delivered dose was estimated by overlaying the planned dose distribution on each day MVCT according to the daily positional shift, and by recalculating the dose distribution based on the treatment sinogram. Accumulated and projected dose of each day were also generated using these two methods for comparison. For a particular fraction, accumulated dose was generated by summing all available daily doses using DIR. For projected dose, daily delivered dose of that fraction was being projected for the remaining fractions and summed to give the accumulated dose. Same DIR map was used in both dose estimation methods. Selected dose endpoints of the targets and critical structures were used for comparison. Results: The mean endpoint doses estimated by dose shifting method were generally lower than those by recalculation method, with the dose deviation for primary GTV, lymph node (LN) PTVs, parotids and spinal cord ranged from 3.2% to 1.8%. For the primary PTV and brainstem, their mean D98 and Dmax were 3.8 4.4% and 5.9 10.5% lower in daily dose, 3.9 3.2% and 7.1 8.1% lower in accumulated dose, and 4.6 7.4% and 5.2 10.6% lower in projected dose using dose shifting method compared to recalculation method respectively. With dose shifting method, only 84%, 63%, 92%, 87%, 79%, 72%, 32% and 79% fractions would have the endpoint doses < 5% dose difference from the recalculation method for primary GTV and PTV, left and right LN PTVs, left and right parotids, brainstem and spinal cord respectively. These values dropped to 62%, 47%, 78%, 69%, 61%, 53%, 5% and 67% if dose acceptance level was set on 3%. Among 3 dose assessment approaches, the dose difference between the two methods was the smallest in accumulated dose and greatest in daily dose. Conclusions: Allowing 5% dose difference, planned dose shifting method provided a fair estimation for dose evaluation during H&N ART. However, one should be noted the dose deviation could be dramatic for structures close to the dose fall off region such as brainstem. Daily dose estimated by recalculation should be used whenever available. Author Disclosure: W. Fung: None. G. Chiu: None. L. Lee: None.