Rationale and Objectives: Recent brain network studies have been broadly applied in Parkinson's disease (PD), but the relationship between structural and metabolic networks remains unclear.This study aimed to use integrated PET/MRI to investigate whether decoupling exists between structural and metabolic networks in the brains of PD patients, providing imaging evidence to elucidate the mechanisms of network disruption in PD. Materials and Methods:We collected diffusion tensor imaging (DTI) data and 18 F-FDG PET imaging data and from 43 patients with PD and 25 healthy controls (HC), generated structural connectivity (SC) and metabolic connectivity (MC) networks.Graph-theoretical methods were used to characterize the topological properties of networks.Based on Yeo 7-network parcellation, whole brain was divided into 7 functional modules, and the structural-metabolic connectivity (SC-MC) coupling was compared at both regional and module level.Results: Both SC and MC networks of PD patients retained small-world properties.For global topological properties, MC network showed longer characteristic path length (Lp) and lower global efficiency (Eg), whereas SC network did not differ from controls.At whole-brain level, compared with HC, significant SC-MC decoupling in bilateral orbital inferior frontal gyrus (IFGorb), bilateral orbital middle frontal gyrus (MFGorb) and right pallidum (PG) (all with p FDR < 0.001) was showed in PD patients.At module level, stronger SC-MC coupling of somatomotor network (SMN) (p FDR = 0.04) was found in PD patients.However, visual network (VIS) (p FDR < 0.001), ventral attention network (VAN) (p FDR = 0.019) and frontoparietal control network (FPCN) (p FDR = 0.027) showed significant SC-MC decoupling in PD patients. Conclusion:Our results showed that PD is associated with reduced metabolic network integration and regional SC-MC decoupling, with the most prominent effects located in orbitofrontal-basal ganglia circuits.SC-MC may provide information complementary to SCfunctional connectivity (FC) and shows promise as a multimodal, network-level biomarker.
Wang et al. (Wed,) studied this question.