目的：在清華水池式反應器(THOR)特定照射條件下，本研究利用特定生物方法來探討THOR做為硼中子捕獲治療(BNCT)用途之超熱中子束的RBE值。材料與方法：本研究所採用之生物方法為小鼠體內腸小囊的再生細胞，實驗過程小鼠並未注射含硼藥物，實驗所得之RBE值則是比較鈷60照射及THOR照射條件下（由不同加馬及“非加馬”輻射劑量組成），產生相同輻射生物效應之輻射劑量的關係值。結果：本研究在THOR兩種照射條件下，超熱中子束的RBE值分別為1.5和1.9，其中加馬輻射劑量所佔的比例分別為75%和70%。假設加馬輻射的RBE值等於1，則THOR兩種照射條件下之“非加馬”輻射的RBE值可分別求得為2.9和4.0。比較THOR與全世界其他6個BNCT設施在相同實驗方法下所得的結果可以發現：即使在設施組成及輻射劑量組合相類似的情況下，各超熱中子束的RBE值亦不盡相同；然而，中子束RBE值隨著“非加馬”輻射劑量所佔的比例增加而微幅增加的趨勢則是一致。雖然各實驗設施所推得之“非加馬”輻射的RBE值亦呈現極大差異，但其平均值3.4則與現行臨床應用來決定總生物權重劑量的RBE值相近。討論與結論：比較THOR與世界上其他BNCT設施的結果顯示：所有設施之超熱中子束的RBE值皆隨著“非加馬”輻射劑量所佔的比例增加而增加，然而，不同設施之間所得的結果差異頗大，無法以設施組成不同或所提供的劑量率（反應器功率）不同來加以解釋，可能要從減少輻射劑量測量的不準度，並採取共通的劑量測量程序，以確保輻射生物數據的一致性。由於本實驗方法具有可全面檢查照射程序（包含劑量）的性質，因此可用來做為新設施建立或重大設施升等的評估依據。

Purpose: To determine the RBE of the BNCT epithermal neutron beam produced at THOR for a reference biological system and reference irradiation conditions. Materials and Methods: Intestinal crypt regeneration in mice was chosen as biological system. No boron compound was administrated. RBE was determined relative to cobalt-60 gamma rays for two irradiation conditions corresponding to different gamma and "non-gamma" absorbed dose mixtures. Results: The RBE(subscript beam) (i.e. considering the beam "as it is", with all dose components and its actual dose rate) was found equal to 1.5 and 1.9 for absorbed dose mixtures comprising 75 and 70% of gamma respectively. Assuming that the RBE of the gamma dose component is equal to unity and deriving the RBE of the non-gamma dose component, RBE(subscript non-γ) values of 2.9 and 4.0 were obtained from the data of the first and second absorbed dose mixture respectively. Comparing the THOR data with those obtained from the same system in 6 other BNCT facilities worldwide, large differences between the RBE(subscript beam) are observed, even when comparing those beam configurations exhibiting similar absorbed dose mixtures. There is, however, a general trend for the RBE(subscript beam) to increase slightly with increasing non-gamma dose component contribution. The derived RBE(subscript non-γ) exhibit also large differences, but average 3.4, which value is close to the one currently apply in clinical practice for determining the "total biologically weighted dose". Discussion and Conclusion: The THOR data as well as those of the other facilities exhibit a certain degree of internal consistency in that the RBE(subscript beam) increases with the non-gamma dose component. However, the RBE differences between institutions appear too large to be explained only by the difference in the composition of the beams and in dose rate (power of the reactor). It is likely that lessening the dosimetric uncertainties and adopting common dosimetric protocols would restore some homogeneity in the radiobiological data. As such experiments constitue an overall check of the irradiation procedure (including dosimetry), they should be undertaken for the commissioning of new installations or on the occasion of any substantial upgrade.