用于对通过氟-18-氟脱氧葡萄糖获得的小鼠PET图像定量血液时间-活性曲线的非侵入性方法。

Noninvasive methods for quantitating blood time-activity curves from mouse PET images obtained with fluorine-18-fluorodeoxyglucose.

作者信息

Green L A, Gambhir S S, Srinivasan A, Banerjee P K, Hoh C K, Cherry S R, Sharfstein S, Barrio J R, Herschman H R, Phelps M E

机构信息

Crump Institute for Biological Imaging, Department of Molecular and Medical Pharmacology, University of California at Los Angeles School of Medicine, 90095-1770, USA.

出版信息

J Nucl Med. 1998 Apr;39(4):729-34.

PMID:9544690

Abstract

UNLABELLED

The mouse model is currently being explored for various applications with PET imaging. Low resolution of current animal scanners relative to mouse size leads to difficulty in quantitating data from mouse PET images. We have, therefore, investigated methods for determining blood time-activity curves (TACs) from mouse PET studies done with fluorine-18-fluorodeoxyglucose (FDG).

METHODS

Eight mice were fasted, the tail vein was injected with 150-300 microCi of FDG and dynamic images were acquired with a CTI/Siemens (Knoxville, TN) animal tomograph for 64.5 min. Concurrently, 11-14 left ventricle (LV) blood samples were drawn directly from the LV chamber. Organ TACs were obtained by drawing circular regions of interest (ROIs) of various sizes on images of the heart, liver and brain. For each mouse, the FDG model parameter K = (K1 x k3)/(k2 + k3) was estimated by a Patlak algorithm with various estimates of the blood TAC and, as a reference tissue TAC, the brain TAC.

RESULTS

Most partial-volume-corrected heart ROI TACs overestimated the LV samples. Blood TACs from heart images produced statistically different estimates of K than did the LV samples. The liver image-derived blood TACs yielded estimates of K that were comparable to those yielded by the LV samples. Estimates of K determined with two directly sampled LV points in conjunction with the liver image-derived TAC were not statistically different from the estimates obtained with the LV samples. The size and location of ROIs on images of the liver minimally affected the TACs.

CONCLUSION

We have shown that it is experimentally possible to obtain a blood TAC from mouse studies by repeatedly sampling from the LV. We have also shown that images of the liver can be used to reliably estimate the blood TAC. Future FDG PET studies with the mouse model will benefit from this demonstrated ability to noninvasively quantitate blood TACs directly from FDG PET images.

摘要

未标记

目前正在探索将小鼠模型用于正电子发射断层扫描（PET）成像的各种应用。当前动物扫描仪相对于小鼠大小的低分辨率导致难以对小鼠PET图像数据进行定量分析。因此，我们研究了从使用氟 - 18 - 氟脱氧葡萄糖（FDG）进行的小鼠PET研究中确定血液时间 - 活性曲线（TAC）的方法。

方法

8只小鼠禁食，经尾静脉注射150 - 300微居里的FDG，并用CTI/西门子（田纳西州诺克斯维尔）动物断层扫描仪采集64.5分钟的动态图像。同时，直接从左心室腔抽取11 - 14份左心室（LV）血液样本。通过在心脏、肝脏和脑图像上绘制不同大小的圆形感兴趣区域（ROI）来获得器官TAC。对于每只小鼠，使用Patlak算法，通过对血液TAC的各种估计以及作为参考组织TAC的脑TAC来估计FDG模型参数K =（K1×k3）/（k2 + k3）。

结果

大多数经部分容积校正的心脏ROI TAC高估了LV样本。来自心脏图像的血液TAC对K的估计与LV样本相比在统计学上有差异。来自肝脏图像的血液TAC对K的估计与LV样本产生的估计相当。用两个直接采样的LV点结合来自肝脏图像的TAC确定的K估计值与用LV样本获得的估计值在统计学上无差异。肝脏图像上ROI的大小和位置对TAC的影响最小。