Poli G L, Bianchi C, Guerra U P
Department of Health Physics, Ospedali Riuniti di Bergamo, Bergamo, Italy -
Q J Nucl Med Mol Imaging. 2013 Dec;57(4):391-400. Epub 2013 Jun 11.
The BasGan algorithm has been specifically developed for semi-automatic quantification of [123I]FP-CIT SPECT studies. The aim of this work was to evaluate the software linearity and its dependence on the main acquisition parameters of the SPECT study.
The anthropomorphic Striatal Phantom filled with different striatal to background activity concentration ratios was used to verify software linearity. The software response was studied with different acquisition parameters and system configurations. For this purpose, phantom studies were performed with varying radius of rotation, pixel size, number of projections and with different collimator types. For each configuration the tomographic spatial resolution was also determined using the Triple Line Source Phantom. The reconstructed SPECT images were corrected for attenuation and scatter.
The BasGan outcome highly correlates with the real Specific Binding Ratio when fixed gamma camera configuration and reconstruction method are used. However, the software quantification depends on system configuration and decreases with increasing partial volume effect. A comparison of the BasGan output with tomographic spatial resolution data for each configuration shows that the software response highly correlates with this quantity.
The BasGan algorithm is a free, easy-to-use and solid tool for quantification of [123I]FP-CIT images. The software outcome depends on system configuration, but highly correlates with tomographic spatial resolution. Then, the measurement of this quantity turns out to be a simple method to normalize the BasGan quantification for different imaging devices, making it possible the use of available reference values for normal subjects and multicentre studies.
BasGan算法是专门为[123I]FP-CIT单光子发射计算机断层扫描（SPECT）研究的半自动定量分析而开发的。本研究的目的是评估该软件的线性度及其对SPECT研究主要采集参数的依赖性。
使用填充有不同纹状体与背景活性浓度比的拟人化纹状体模体来验证软件的线性度。研究了该软件在不同采集参数和系统配置下的响应。为此，进行了模体研究，改变了旋转半径、像素大小、投影数量，并使用了不同的准直器类型。对于每种配置，还使用三线源模体确定了断层空间分辨率。对重建的SPECT图像进行了衰减和散射校正。
当使用固定的γ相机配置和重建方法时，BasGan的结果与实际特异性结合率高度相关。然而，软件定量分析取决于系统配置，并随着部分容积效应的增加而降低。将每种配置下BasGan的输出与断层空间分辨率数据进行比较，结果表明软件响应与该量高度相关。
BasGan算法是一种免费、易于使用且可靠的[123I]FP-CIT图像定量分析工具。软件结果取决于系统配置，但与断层空间分辨率高度相关。因此，测量该量成为一种简单的方法，可针对不同成像设备对BasGan定量分析进行标准化，从而使得能够使用正常受试者的可用参考值并开展多中心研究。
