Differences between the image-derived input functions of the descending aorta and carotid artery in brain pharmacokinetic parametric analysis

作者：Zixiang Chen, Yaping Wu, Bolun Li, Zhi-Hao Che, Yuxi Jin, Lingxin Chen, Yongfeng Yang, Hairong Zheng, Dong Liang, Meiyun Wang, Zhanli Hu · 发表于：British Journal of Radiology · 年份：2025 · DOI：10.1093/bjr/tqaf166 · 研究领域：Medical Imaging Techniques and Applications、Cerebrovascular and Carotid Artery Diseases、Radiation Dose and Imaging

OBJECTIVES: This study aimed to quantitatively reveal the difference between the descending aorta image-derived input function (IDIFa) and carotid artery image-derived input function (IDIFc) in the task of brain pharmacokinetic parametric analysis. METHODS: 18F-fluorodeoxyglucose (18F-FDG) dynamic positron emission tomography (PET) data from 50 patients collected by UIH uEXPLORER PET/CT were used to extract the IDIFa and IDIFc; the Patlak graphical model (PGM) and irreversible 2-tissue compartmental model (2TCM) were applied for brain pharmacokinetic parametric analysis. The Patlak Ki and Vt images yielded by the 2 types of IDIFs are compared using relative error (RE), while the parameters K1∼k3 and fv were analysed via paired t-tests to determine the significance of the differences. RESULTS: IDIFa and IDIFc had significant differences in peak value (P = 0.0009) and curve integral value (P = 0.0012). Based on PGM, 0%-15% REs were found between the Ki and Vt images yielded by IDIFa and IDIFc; based on 2TCM, significant differences were found in brainstem's K1 (P = 0.0009), left cerebellum's k3 (P = 0.0009) and all selected regions' fv and Ki (P < 0.05). Fitting residuals of IDIFa are lower than that of IDIFc without significant differences (P > 0.05). CONCLUSIONS: IDIFa and IDIFc have different curves characteristics and different brain pharmacokinetic parametric analysis results based on PGM and 2TCM. One should be prudential when using IDIFa to conduct brain dynamic PET para...