Oh Seungjong, Awan Musaddiq J, Monroe James I, Liang Yun, Wegner Rodney E, Karlovits Stephen, Machtay Mitchell, Lo Simon S, Sloan Andrew, Sohn Jason W
Division of Radiation Oncology, Allegheny Health Network Cancer Institute, Pittsburgh, Pennsylvania, USA.
Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, Wisconsin, USA.
Med Phys. 2022 May;49(5):2931-2937. doi: 10.1002/mp.15619. Epub 2022 Apr 12.
To develop a volume-independent conformity metric called the Gaussian Weighted Conformity Index (GWCI) to evaluate stereotactic radiosurgery/radiotherapy (SRS/SRT) plans for small brain tumors.
A signed bi-directional local distance (BLD) between the prescription isodose line and the target contour is determined for each point along the tumor contour (positive distance represents under-coverage). A similarity score function (SF) is derived from Gaussian function, penalizing under- and over-coverage at each point by assigning standard deviations of the Gaussian function. Each point along the dose line contour is scored with this SF. The average of the similarity scores determines the GWCI. A total of 40 targets from 18 patients who received Gamma-Knife SRS/SRT treatments were analyzed to determine appropriate penalty criteria. The resulting GWCIs for test cases already deemed clinically acceptable are presented and compared to the same cases scored with the New Conformity Index to determine the influence of tumor volumes on the two conformity indices (CIs).
A total of four penalty combinations were tested based on the signed BLDs from the 40 targets. A GWCI of 0.9 is proposed as a cutoff for plan acceptability. The GWCI exhibits no target volume dependency as designed.
A limitation of current CIs, volume dependency, becomes apparent when applied to SRS/SRT plans. The GWCI appears to be a more robust index, which penalizes over- and under-coverage of tumors and is not skewed by the tumor volume.
开发一种名为高斯加权适形指数（GWCI）的与体积无关的适形度量指标，以评估小脑肿瘤的立体定向放射外科/放射治疗（SRS/SRT）计划。
确定处方等剂量线与肿瘤轮廓上各点之间的有符号双向局部距离（BLD）（正距离表示覆盖不足）。从高斯函数导出相似性评分函数（SF），通过指定高斯函数的标准差对各点的覆盖不足和过度覆盖进行惩罚。剂量线轮廓上的每个点都用此SF进行评分。相似性评分的平均值确定GWCI。对18例接受伽玛刀SRS/SRT治疗的患者的40个靶区进行分析，以确定合适的惩罚标准。给出了已被临床认可的测试病例的GWCI结果，并与用新适形指数评分的相同病例进行比较，以确定肿瘤体积对两种适形指数（CI）的影响。
基于40个靶区的有符号BLD，共测试了四种惩罚组合。建议将GWCI为0.9作为计划可接受性的临界值。GWCI按设计不显示靶区体积依赖性。
当前CI的一个局限性，即体积依赖性，在应用于SRS/SRT计划时变得明显。GWCI似乎是一个更稳健的指数，它对肿瘤的覆盖不足和过度覆盖进行惩罚，且不受肿瘤体积的影响。
