且的：研究螺旋斷層加速器放射治療，影響擺位誤差的治療相關因素。在臨床治療因擺位誤差， 並建議計畫靶體積（planning target volume, PTV)所需增加之安全邊綠值。 材料與方法:文獻指出可能影響病人治療擺位誤差的相關因素，包括系統誤差（S)和隨機誤 差（〇)。影響〇的因素，包括病人年齡，性別，腫瘤位置，身體質量指數（body mass index, BMI)，體重減輕，腫瘤分期，及腫瘤轉移等因素。本研究收集台灣南部某區域醫院放射治療部 門2013至2016年，共138位癌症病人接受螺旋斷層加速器治療。治療前以百萬伏特（megavoltage)電腦斷層攝影定位，執行影像導引放射治療（image-guided radiation therapy, IGRT) 。 總共收集治療期間各分次治療4081組影像進行擺位誤差分析。計算每位病人治療期間X, Y, Z三 軸位移，三維空間（3D)向量，及Y軸（roll)旋轉的擺位誤差平均值，及標準差。利用文獻公式 計算得出S和^數值。分析影響病人治療X, Y, Z三軸，3D向量，及roll旋轉的擺位誤差的因素。 參考van Herk學者提出公式（2.5S+0.7g)演算，得出於臨床治療PTV所需建議增加安全邊綠值。 結果：研究治療期間影響X，Y，Z三軸，3D向量，及roll旋轉的擺位誤差平均位移值，發現與 病人年齡（p= 0.05)，腫瘤位置（p< 0.001 )，及BMI (p< 0.001 )，達統計顯著差異性。單變 項迴歸分析影響3D向量擺位誤差的因素，發現腫瘤位置，體重減輕，腫瘤期別，達統計顯著 差異性。另單變項迴歸分析影響roll旋轉的擺位誤差的因素，顯示只有腫瘤位置影響擺位誤差因 素，達統計顯著差異性。經多變項迴歸逐步篩選分析，發現只有腫瘤位置達統計顯著差異性（p< 0.001 )。經計算PTV所需建議增加X，Y，Z三軸安全邊綠值，在胸部，腹部，骨盆部和前列腺 腫瘤需要5-11 mm，而70-89歲年齡層及BMI 225安全邊綠建議值分別為6-8 mm及5-6 mm。 結論：本研究發現螺旋斷層加速器治療期間影響擺位誤差的重要因素為病人年齡、腫瘤位置及 BM卜PTV安全邊綠建議值，需針對這些影響擺位誤差因素，有不同的治療安全邊綠建議值。提 供做為螺旋斷層治療器的臨床治療的參考。

Objective : We investigate treatment-related factors to affect setup errors during the course in radiotherapy using helical TomoTherapy® and recommend required safety margins for planning target volume (PTV) clinically. Materials and Methods : The literature indicates that may affect setup errors including systematic error (S) and random error (a). The factors to impact a in radiotherapy may be caused by age of patient, gender, tumor site, body mass index (BMI), weight loss, cancer stage, and the status whether metastasis. In the study, we collected a total of 138 patients who received helical TomoTherapy® from 2013 to 2016 at a regional hospital in southern Taiwan. All patients received megavoltage computed tomography for daily therapy position checkup. A total of 4081 sets of images were collected for analysis of setup error. We calculate means, standard deviations of daily setup error of each patient in three axis directions (X, Y, Z-axis), three dimensional (3D) vector, and Y-axis (roll) rotation during the treatment. Then S and a are derived from adopted equations in the literature. In the study we analyzed which to factors to influence setup error during treatment. The formula (2.5S+0.7a) by the author van Herk et al. is proposed to recommend safety margin of PTV clinically. Results : The factors significantly affect daily setup error in X, Y, Z-axis, 3D vector, and Y-axis (roll) rotation were in groups of patienfs age (p= 0.05), tumor sites (p< 0.001), and BMI (p< 0.001). Single regression analysis in factors to affect 3D vector setup error, we found tumor sites, weight loss, and tumor stage to achieve statistical significance. Another single regression analysis in factors to affect Y-axis (roll) rotation setup error, we disclosed only tumor site to achieve statistical significance. After multiple regression model, we concluded tumor site is the only factor to influence setup error significantly (p< 0.001). In the study, we recommended wider safety margins of PTV are 5-11 mm in X, Y, Z- axes in chest, abdomen, pelvis, and prostate tumors. The safety margins are 5-8 mm and 5-6 mm in the age group of 70-89 and the BMI >25 respectively. Conclusion : The important factors to affect setup error are age of patient, tumor site, and BMI. In this study, we recommend the safety margin in PTV should be individualized based on these setup error factors in the clinical practice of helical TomoTherapy®.