立體定位軀體放射治療技術(Stereotactic body radiation therapy, SBRT)，其特色為針對較小體積腫瘤，以少治療次數、單次治療劑量高於傳統放射治療3 至15 倍以上施行，因此臨床上必須結合影像導引(Image guidance)執行，以提高治療準確度。本研究選取肺腫瘤及骨轉移14 名病人，共54 次治療，依不同日常體能狀態(Performance status, PS)、不同類別固定輔具作分類，使用醫用直線加速器(Elekta Synergy® , UK)之錐狀射束電腦斷層影像系統(cone-beam CT)於不同時間(擺位後、位移校正、開始治療起每間隔十分鐘、治療結束)進行影像比對驗證位移誤差。研究結果顯示，僅在治療時間超過20 分鐘後，Z 軸有統計上顯著差異(P<0.05)，但誤差值皆小於1mm，不具臨床意義。依量身設置的四種輔具：(1)僅使用真空固定墊、(2)真空固定墊與膝部固定墊、(3)真空固定墊與腹部壓迫板、(4)真空固定墊、膝部固定墊加腹部壓迫板，及不同病人日常體能狀態(PS 值介於0 到2)，其分析顯示兩項因子與治療中位移誤差均無統計上明顯差異(P>0.05)，其可能原因為選取病人之體能狀態大多介於0 到1 之間，搭配適合固定輔具使用，大幅增加治療位置穩定度及再現性。因此，建議執行立體定位軀體放射治療，對於低疼痛指數及日常體能狀態指數小於1 之病人，於治療期間每間隔10 分鐘執行影像導引驗證，即足以確保治療期間內之位移準確性。

Stereotactic body radiation therapy (SBRT) is characterized by using fewer treatment fractions and higher fraction size (3 to 15 times more intense than conventional radiation) for small tumor. Therefore it is very important to incorporate the image guidance procedure to increase the accuracy of the treatment delivery. This study enrolled 14 patients
undergoing SBRT for lung tumors and bone metastases. The performance status (PS) and immobilization method were recorded. We use the linear accelerator (Elekta Synergy® , UK) with on board cone-beam CT to ensure the patient setup error at various time point (localization、verification、intrafraction、post-treatment ). The result showed a significant
difference in the Z-axis 20 minutes after starting SBRT (P<0.05). However, the error is less than 1mm though, and is not considered to be clinically significant. We use 4 different immobilization methods, including vacuum bag alone, vacuum bag plus knee support, vacuum bag plus abdominal compression plate, and vacuum bag plus abdominal compression plate
and knee support. The analyses revealed that the intrafraction setup error was not significant associated with different immobilization methods (P>0.05) or different PS (P>0.05), possibly due to most patients have performance status of 0-1 with the appropriate immobilization to increase stability and reproducibility. In conclusion, for patients with low pain
score, performance status <1, and suitable immobilization, imaging by cone-beam CT with every 10-minutes interval is more than adequate to ensure accuracy during SBRT treatment.