Category: Parkinson's Disease: Neuroimaging
Objective: To create a voxel-based method, oriented to the brainstem and allowing the detection of Parkinson’s disease (PD) and the study of the altered connectivity
Background: Due to the smallness of the brainstem nuclei and the uncertainties regarding their exact locations and the functional disturbances induces by PD, we sought to develop an automatic voxel-based analysis by comparing a population of PD patients to controls. We chose beforehand to register the patient on a model, using for registration the simultaneously acquired NMR T1 images and then applying the same transformations to the PET images.
Method: 51 PD patients and 72 controls underwent, on a PET/NMR hybrid system, an inversion-recovery T1W MRI acquisition followed by a 18F-FDOPA PET acquisition. After registration of both 3D-images on a brainstem model and reduction of the nonspecific activity, the SUV image of the 18F-FDOPA of the whole brain was converted to the Z-scale. A probability value was assigned to each voxel, using the statistically normal distribution of controls and taking into account the Z-score of its neighbours. Only the voxels with a significantly abnormal value were further considered. They were grouped, based on intercorrelation and/or on raw anatomical delimitation. The final processing included distant crosscorrelation for the study of the connectivity and Principal Component and Factor Discriminant Analyses for the PD prediction.
Results: Focal distortions were induced by the registration in 12 of the 123 patients, but they were almost exclusively located in the cortex and without severe impact. In PD, pathological areas were observed in the striatum, caudate, substantia nigra, accumbens n., pedunculopontine n., reticular formation of the medulla oblongata, amygdala, thalamus, cingulum and to a lesser extent he olfactory bulb and the parahippocampal gyrus. The overall accuracy of PD detection was 98%. The connectivity inside the brainstem reveals a right dominancy on controls, disappearing in PD patients.
Conclusion: Although retrospective and monocentric, this study reveals the feasibility to conduct a voxel-based analysis on the brainstem. In presence of a hybrid system (PET/NMR or probably PET/CT), the methodology has a wide scope of application: it can easily and robustly be applied to other PET tracers.
To cite this abstract in AMA style:Q. Demonceau, G. Demonceau, V. Lebon. A brainstem-oriented PET/NMR method in Parkinson’s disease [abstract]. Mov Disord. 2022; 37 (suppl 2). https://www.mdsabstracts.org/abstract/a-brainstem-oriented-pet-nmr-method-in-parkinsons-disease/. Accessed March 2, 2024.
« Back to 2022 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/a-brainstem-oriented-pet-nmr-method-in-parkinsons-disease/