Calvary Mater Newcastle Hospital, Newcastle, New South Wales, 2298, Australia.
Med Phys. 2009 Dec;36(12):5488-96. doi: 10.1118/1.3253463.
The aim of this work was to investigate the use of amorphous silicon electronic portal imaging devices (EPIDs) for regular quality assurance of linear accelerator asymmetric jaw junctioning.
The method uses the beam central axis position on the EPID measured to subpixel accuracy found from two EPID images with 180 degrees opposing collimator angles. Individual zero jaw position ("half-beam blocked") images are then acquired and the jaw position precisely determined for each using penumbra interpolation. The accuracy of determining jaw position with the EPID method was measured by translating a block (simulating a jaw) by known distances, using a translation stage, and then measuring each translation distance with the EPID. To establish the utility of EPID based junction dose measurements, radiographic film measurements of junction dose maxima/minima as a function of jaw gap/overlap were made and compared to EPID measurements. Using the method, the long-term stability of zero jaw positioning was assessed for four linear accelerators over a 1-1.5 yr time period. The stability at nonzero gantry angles was assessed over a shorter time period.
The accuracy of determining jaw translations with the method was within 0.14 mm found using the translation stage [standard deviation (SD) of 0.037 mm]. The junction doses measured with the EPID were different from film due to the nonwater equivalent EPID scattering properties and hence different penumbra profile. The doses were approximately linear with gap or overlap, and a correction factor was derived to convert EPID measured junction dose to film measured equivalent. Over a 1 yr period, the zero jaw positions at gantry zero position were highly reproducible with an average SD of 0.07 mm for the 16 collimator jaws examined. However, the average jaw positions ranged from -0.7 to 0.9 mm relative to central axis for the different jaws. The zero jaw position was also reproducible at gantry 90 degrees position with 0.1 mm SD variation with the mean jaw position offset from the gantry zero position consistently by 0.3-0.4 mm for the jaws studied.
The EPID based method is efficient and yields more precise data on linear accelerator jaw positioning and reproducibility than previous methods. The results highlight that zero jaw positions are highly reproducible to a level much smaller than the displayed jaw resolution and that there is a need for better methods to calibrate the jaw positioning.
本研究旨在探讨使用非晶硅电子射野影像装置（EPID）对直线加速器不对称准直器连接进行常规质量保证。
该方法利用从两个具有 180 度相反准直器角度的 EPID 图像中以亚像素精度找到的 EPID 上的射束中心轴位置。然后获取单个零准直器位置（“半束阻挡”）图像，并使用半影插值精确确定每个准直器位置。使用平移台以已知距离平移块（模拟准直器），然后使用 EPID 测量每个平移距离，从而测量 EPID 确定准直器位置的准确性。为了建立基于 EPID 的连接剂量测量的实用性，对作为准直器间隙/重叠函数的连接剂量最大值/最小值进行了射线照相胶片测量，并与 EPID 测量进行了比较。使用该方法，在 1-1.5 年的时间内评估了四台直线加速器的零准直器定位的长期稳定性。在较短的时间内评估了非零旋转架角度的稳定性。
使用平移台确定准直器平移的准确性在 0.14 毫米以内[标准偏差（SD）为 0.037 毫米]。由于 EPID 的非等效散射特性以及不同的半影轮廓，EPID 测量的连接剂量与胶片不同。剂量与间隙或重叠大致呈线性关系，并且推导了一个校正因子，将 EPID 测量的连接剂量转换为胶片测量的等效剂量。在 1 年的时间内，在旋转架零度位置，零准直器位置具有高度可重复性，16 个准直器的平均 SD 为 0.07 毫米。然而，对于不同的准直器，准直器位置的平均值在-0.7 到 0.9 毫米之间，相对于中心轴。在旋转架 90 度位置，零准直器位置也具有可重复性，16 个准直器的平均 SD 变化为 0.1 毫米，平均准直器位置相对于旋转架零位的偏移始终为 0.3-0.4 毫米。
基于 EPID 的方法高效，并且比以前的方法提供了更精确的直线加速器准直器定位和可重复性数据。结果表明，零准直器位置具有高度的可重复性，达到比显示的准直器分辨率小得多的水平，并且需要更好的方法来校准准直器位置。
