Glucose metabolism images of normal subjects acquired with FDG-PET, dependence on data acquisition and its correction.

正电子发射断层扫描（PET）与F-18标记的氟脱氧葡萄糖（FDG）提供了非侵入性成像的葡萄糖代谢的人脑，是有用的诊断阿尔茨海默氏病，癫痫和其他神经系统疾病。然而，在每个PET中心收集足够数量的正常数据并不容易，并且在其他中心采集的数据可能无法使用。五个PET中心参与了这项研究项目，以创建一个正常的数据库，以促进FDG-PET测量的标准化。将数据转换为Dr. View或Analyze格式。当FDG动脉浓度的时间过程被归一化的注射剂量每体表面积，这已被描述为最佳的身高和体重，输入函数的积分导致在7%的个体变异。在标准化输入函数中未观察到中心间差异。考虑到噪声，注射后45-60分钟测量的局部脑FDG摄取除以积分输入几乎（尽管不完全）等于用动力学分析估计的FDG流入速率常数，并提供了对葡萄糖代谢速率的良好估计。结合输入函数的标准化，这表明了从单帧PET图像估计脑葡萄糖代谢率而无需动脉采样的可能性。当用SPM对老年人的FDG图像进行空间标准化时，在某些情况下会错误地确定屈伸角。当使用CBF模板图像用SPM对FDG图像进行空间归一化时，枕叶被扭曲并向下推入小脑，这表明有必要使用FDG模板图像进行空间归一化。

Positron emission tomography (PET) with F-18 labeled fluorodeoxyglucose (FDG) provides noninvasive imaging of glucose metabolism of the human brain and is useful for diagnosis of Alzheimer's disease, epilepsy and other neurological disorders. However, it is not easy to collect a sufficient number of normal data in each PET center, and the data acquired in other centers may not be used. Five PET centers participated in this research project to create a normal database to facilitate standardization of FDG-PET measurement. Data were transformed into Dr. View or Analyze format. When the time course of FDG arterial concentration was normalized by the injected dose per body surface area, which had been described optimally with height and body weight, the integral of the input function resulted in an individual variation of 7%. Between-center difference was not observed in the standardized input function. Allowing for the noise, the regional cerebral FDG uptake measured 45-60 minutes post injection divided by the integrated input was almost, though not exactly, equal to the influx rate constant of FDG estimated with a kinetic analysis, and provided a good estimate of the rate of glucose metabolism. Combined with standardization of the input function, this suggested a possibility of estimating the cerebral metabolic rate of glucose from a single-frame PET image without arterial sampling. When the FDG images of elderly subjects were spatially normalized with SPM, the flexion-extension angle was erroneously determined in some cases. When the FDG images were spatially normalized with SPM using the CBF template image, the occipital lobe was distorted and pushed downward into the cerebellum, suggesting the necessity of using an FDG template image for spatial normalization.