Department of Computer Science and Information Engineering, National Dong Hwa University, Hualien 97401, Taiwan.
Ann Nucl Med. 2011 Dec;25(10):796-803. doi: 10.1007/s12149-011-0533-8. Epub 2011 Sep 2.
6-[(18)F]Fluoro-L: -DOPA (FDOPA) is a radiopharmaceutical valuable for assessing the presynaptic dopaminergic function when used with positron emission tomography (PET). More specifically, the striatal-to-occipital ratio (SOR) of FDOPA uptake images has been extensively used as a quantitative parameter in these PET studies. Our aim was to develop an easy, automated method capable of performing objective analysis of SOR in FDOPA PET images of Parkinson's disease (PD) patients.
Brain images from FDOPA PET studies of 21 patients with PD and 6 healthy subjects were included in our automated striatal analyses. Images of each individual were spatially normalized into an FDOPA template. Subsequently, the image slice with the highest level of basal ganglia activity was chosen among the series of normalized images. Also, the immediate preceding and following slices of the chosen image were then selected. Finally, the summation of these three images was used to quantify and calculate the SOR values. The results obtained by automated analysis were compared with manual analysis by a trained and experienced image processing technologist.
The SOR values obtained from the automated analysis had a good agreement and high correlation with manual analysis. The differences in caudate, putamen, and striatum were -0.023, -0.029, and -0.025, respectively; correlation coefficients 0.961, 0.957, and 0.972, respectively.
We have successfully developed a method for automated striatal uptake analysis of FDOPA PET images. There was no significant difference between the SOR values obtained from this method and using manual analysis. Yet it is an unbiased time-saving and cost-effective program and easy to implement on a personal computer.
6-[(18)F]氟-L-多巴（FDOPA）是一种放射性药物，当与正电子发射断层扫描（PET）一起使用时，可用于评估突触前多巴胺能功能。更具体地说，FDOPA 摄取图像的纹状体与枕叶的比值（SOR）已被广泛用作这些 PET 研究中的定量参数。我们的目的是开发一种能够对帕金森病（PD）患者的 FDOPA PET 图像进行客观分析的简单、自动化方法。
我们的自动纹状体分析纳入了 21 名 PD 患者和 6 名健康受试者的 FDOPA PET 研究的脑图像。对每个个体的图像进行空间归一化到 FDOPA 模板中。随后，在归一化图像系列中选择基底神经节活动水平最高的图像切片。此外，还选择了所选图像的前后两个相邻图像切片。最后，使用这三张图像的总和来定量和计算 SOR 值。通过自动化分析获得的结果与经过培训和经验丰富的图像处理技术人员的手动分析进行了比较。
自动化分析获得的 SOR 值与手动分析具有良好的一致性和高度相关性。尾状核、壳核和纹状体的差异分别为-0.023、-0.029 和-0.025；相关系数分别为 0.961、0.957 和 0.972。
我们已经成功开发了一种用于 FDOPA PET 图像自动纹状体摄取分析的方法。该方法获得的 SOR 值与手动分析无显著差异。然而，它是一种无偏、省时、具有成本效益的程序，易于在个人计算机上实现。
