以單一劑量點膠片利用已知之膠片劑量校正曲線資料建立實際可用之劑量校正曲線。將 EBT2 膠片剪成 2×3公分大小置於固態水假體中照射不同劑量，於一天後用 EPSON 1680 掃描器掃描，以程式取出紅色影像並計算其淨光密度值並建立劑量校正曲線。在不同日期時間建立多組劑量校正曲線後，以其中一組當作未知之劑量曲線取其中一點劑量膠片，把其他多組已知之劑量校正曲線作為參考資料經內插擬合計算後算出此未知組之劑量校正曲線。比較此未知組之劑量校正曲線與計算出之劑量校正曲線分析其劑量差異。由未知組所有劑量點之膠片所擬合之劑量校正曲線與由未知組單點劑量膠片經參考已知組之劑量校正曲線後所計算出之劑量校正曲線之比較，在內插法中，劑量 20-400 cGy 區域間平均劑量百分差異為 1.79±1.04%，而增加到兩點不同劑量量測，可再將低到 0.69±0.32%的差異；在最小距離法中，平均劑量百分差異為 1.78±1.05%，而增加到兩點不同劑量量測，可再將低到0.72±0.31%的差異。本研究之方法經慎選參考劑量點做計算後可獲得實際可用之膠片劑量校正曲線，大幅縮短膠片劑量校正曲線建立之時間。內插法和最小距離法的結果差不多，但已知點為較高劑量時，內插法比最小距離法有更小的差異；已知點為較低劑量時，最小距離法比內插法有更小的差異。

In this study, we established dose calibration curve from film dose database. The EBT2 film was cut into 2×3 cm2
shape and placed in solid water phantom and exposed to a range of irradiation. The film was scanned by using EPSON
1680 scanner after a day, and red channel image was retrieved for the calculation of net optical density for establishing
dose calibration curve. The same procedures were conducted on different film slips on different dates to obtain multiple
dose calibration curves. One calibration curve, namely, test calibration curve (TCC) was randomly selected from the
multiple dose calibration curves, and one data point from the selected curved was random chosen. The selected data point
was used to localize two nearest calibration curves and interpolation, minimum-distance, hybrid methods were used to
find the estimated calibration curves (ECCs), without using other data points measured on the same experiment with the
selected data point. The ECCs identified by the three methods were compared with the TCC. In interpolation method, the
mean difference between ECCs and TCC at 20-400 cGy was 1.79±1.04% when single data points were compared. When
two data points were compared, the mean difference was 0.69±0.32%. In minimum-distance method, the mean difference
was 1.78±1.05% when single data points were compared. When two data points were compared, the mean difference was
0.72±0.31%. We conclude that interpolation and minimum- distance method had similar outcomes. However, when
selected data points were at higher dosage range, the interpolation method had a smaller difference with TCC, comparing
to minimum-distance method. On the contrary, when selected data points were at lower dosage range, the
minimum-distance method had a smaller difference with TCC, comparing to interpolation method. All three methods may
be used as alternates for classic calibration method for saving time and costs consumed in the procedures.