Nishitani Takuya, Matsugi Kiyotomo, Kitano Tetsuya, Yamamoto Hiroyuki, Kaneko Masakazu, Yamada Shigeki
Department of Radiation Oncology, Shiga Medical Center for Adults.
Department of Radiology, Shiga Medical Center for Adults.
Nihon Hoshasen Gijutsu Gakkai Zasshi. 2018;74(1):39-47. doi: 10.6009/jjrt.2018_JSRT_74.1.39.
In radiotherapy planning, CT images are widely used to delineate the gross tumor volume (GTV) and the organs at risks (OARs), which allows for the calculation of the dose distribution to each structure. The delineated contours of the GTV and OARs may become inaccurate, and subsequently result in the inaccurate derivation of the dose distribution, if there are metal artifacts present in the CT image. The metal artifact reduction technique, single energy metal artifact reduction (SEMAR), installed on the CT system (Aquilion ONE Vision Edition, Toshiba Medical Systems Corporation) could potentially reduce metal artifacts. Therefore, we investigated whether SEMAR can improve the accuracy of delineation, and subsequently the dosimetric accuracy, in the treatment planning process. Using an acrylonitrile-butadiene-styrene resin phantom (RT-3000-New, R-Tech. Inc, Tokyo, Japan), titanium bars were inserted on both the left and right sides, and four types of electron density inserts (rods) were separately inserted in the middle. The electron densities of the rods were 0.90, 0.96, 1.07, and 1.09. After CT images were acquired, SEMAR-ON (when applying the SEMAR correction) images were generated. On both SEMAR-ON and SEMAR-OFF (when not applying the SEMAR correction) images, the rod contours were delineated automatically, using a CT value threshold. This threshold was selected so that the area of the automatically delineated contour was 615.4 mm. The difference in the contour area of SEMAR-ON, SEMAR-OFF, and no metal artifact images were compared using the dice coefficient. When SEMAR was used, the dice coefficient improved by 57.4%. Therefore, SEMAR was considered to be useful in improving the accuracy of GTV and OAR delineation.
在放射治疗计划中，CT图像被广泛用于勾画大体肿瘤体积（GTV）和危及器官（OARs），这有助于计算每个结构的剂量分布。如果CT图像中存在金属伪影，GTV和OARs勾画的轮廓可能会不准确，进而导致剂量分布推导不准确。安装在CT系统（Aquilion ONE Vision Edition，东芝医疗系统公司）上的金属伪影减少技术，即单能量金属伪影减少（SEMAR），有可能减少金属伪影。因此，我们研究了SEMAR在治疗计划过程中是否能提高勾画的准确性，进而提高剂量测定的准确性。使用丙烯腈-丁二烯-苯乙烯树脂体模（RT-3000-New，R-Tech. Inc，东京，日本），在左右两侧插入钛棒，并在中间分别插入四种类型的电子密度插入物（棒）。棒的电子密度分别为0.90、0.96、1.07和1.09。采集CT图像后，生成SEMAR开启（应用SEMAR校正时）图像。在SEMAR开启和SEMAR关闭（不应用SEMAR校正时）图像上，使用CT值阈值自动勾画棒的轮廓。选择该阈值以使自动勾画轮廓面积为第615.4平方毫米。使用骰子系数比较SEMAR开启、SEMAR关闭和无金属伪影图像轮廓面积的差异。使用SEMAR时，骰子系数提高了57.4%。因此，SEMAR被认为有助于提高GTV和OAR勾画的准确性。
