Correlation and differential diagnostic value of 18F-FDG imaging and arterial spin labeling cerebral perfusion imaging acquired by synchronous PET/MR in Parkinsonism
10.3760/cma.j.issn.2095-2848.2019.06.004
- VernacularTitle: 帕金森综合征PET/MR同步采集18F－FDG显像与ASL脑血流灌注成像相关性及鉴别诊断价值
- Author:
Xun SUN
1
,
2
;
Fang LIU
1
,
2
;
Weiwei RUAN
1
,
2
;
Xiaojuan HUANG
1
,
2
;
Tao WANG
3
;
Bo HU
3
;
Xiaoli LAN
1
,
2
Author Information
1. Department of Nuclear Medicine, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430022, China
2. Hubei Key Laboratory of Molecular Imaging, Wuhan 430022, China
3. Department of Neurology, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430022, China
- Publication Type:Journal Article
- Keywords:
Parkinsonian disorders;
Positron-emission tomography;
Magnetic resonance imaging;
Diagnosis, differential;
Deoxyglucose
- From:
Chinese Journal of Nuclear Medicine and Molecular Imaging
2019;39(6):337-343
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To discuss the correlation between 18F-fluorodeoxyglucose (FDG) imaging and arterial spin labeling perfusion weighted imaging (ASL-PWI) using synchronous PET/MR, and explore the feasibility of 18F-FDG PET and ASL for differential diagnosis in Parkinsonism (PS).
Methods:A retrospective analysis was performed in 24 patients with Parkinson′s disease (PD), 10 patients with multiple system atrophy (MSA) and 6 patients with progressive supranuclear palsy (PSP) who all underwent 18F-FDG PET/MR imaging from October 2017 to December 2018 (15 males, 25 females, age: 34-77 years). Synchronously acquired data, including MRI T1, 18F-FDG PET and three dimensional (3D) ASL-PWI were extracted. T1 structure image was used to register with standard brain template, and the personalized brain template of each patient was obtained. Then, the templates were matched to the corresponding PET and ASL-PWI images, and the mean standardized uptake value (SUVmean) and mean cerebral blood flow (CBFmean) of each brain region could be extracted. Pearson correlation analysis was used to evaluate the correlation between SUVmean and CBFmean of the whole brain and the regions of interest (ROI; caudate nucleus, putamen, midbrain, pons and cerebellum). One-way analysis of variance and least significant difference t test were used to evaluate the characteristics of SUVmean and CBFmean in different diagnostic groups. The diagnostic values of SUVmean and CBFmean in different ROI were evaluated with receiver operating characteristic (ROC) curve analysis.
Results:A total of 2 800 pairs of SUVmean and CBFmean were obtained from 40 cases in 3 groups. The overall correlation between them was good (r=0.468, P<0.01). There were also significant correlations between SUVmean and CBFmean in caudate nucleus, putamen, pons and cerebellum (r values: 0.346-0.492, all P<0.05). The CBFmean of cerebellum in MSA group was significantly lower than that in PD group ((36.92±11.35) vs (47.92±10.75) ml·100 g-1·min-1; F=4.817, t=2.675, P<0.05). Compared to the PSP group, CBFmean in caudate nucleus was higher in PD group ((30.02±3.51) vs (40.21±8.13) ml·100 g-1·min-1; F=5.393, t=2.969, P<0.05). The similar results were obtained for SUVmean. To differentiate PD with MSA, CBFmean of cerebellum possessed the highest diagnostic efficiency, with the area under ROC curve (AUC), sensitivity and specificity reaching 0.792, 83.3% and 70.0% respectively. To differentiate PSP with PD or MSA, CBFmean in caudate nucleus had the highest diagnostic efficiency, and the AUC were 0.903 and 0.933 respectively.
Conclusions:In patients with PD, MSA and PSP, the level of 18F-FDG glucose metabolism is in good agreement with the ASL cerebral blood flow perfusion obtained by synchronous PET/MR. The distribution characteristics of specific ROI in both PET and MR imaging can be helpful to the differential diagnosis of PS.
