Meige's syndrome (MS), also known as Brueghel's syndrome, is a dystonic disorder predominantly affecting the craniocervical muscles. Characterized by blepharospasm, platysm, and cervical dystonia (Meige, 1910; Pandey & Sharma, 2017) MS is a functionally disabling disease. It causes persistent abnormal facial muscle contractions, leading to complications such as visual impairment, functional blindness, and difficulties in mastication, swallowing, and speech. These symptoms significantly impact patients’ daily lives and quality of life (Peckham et al., 2011). Despite ongoing research, the underlying etiology and pathophysiology of MS remain incompletely understood. Current clinical treatments include botulinum neurotoxin therapy, oral medication, and deep brain stimulation (Fante & Frueh, 2001; Hassell & Charles, 2020; Sáenz-Farret & Zúñiga-Ramírez, 2021; Sako et al., 2011).
Historically, medical imaging has been pivotal in elucidating the pathogenic mechanisms of MS. A study in 2006 was the first to use functional magnetic resonance imaging (fMRI) to highlight impaired motor activation specific to the oral and mandibular motor system in MS patients (Dresel et al., 2006). Recent investigations into MS have increasingly focused on dysfunctions within brain networks. A study by Liu et al. utilized positron emission tomography to examine the metabolic networks in MS patients. This study highlighted that MS patients exhibit abnormal small-world properties in their brain networks, with particular emphasis on the right pallidum and thalamus as critical areas in the development of the disease (Liu et al., 2021). Complementing these findings, another study employed diffusion beam imaging to delve deeper into the brain’s structural connections. This study provides a compelling evidence for a dysfunction in the thalamus-relayed visual and motor network – a central aspect of MS pathology. Concurrently, this research observed an impairment in the microstructural integrity within the dentato-thalamic trajectories. Such findings suggest a potential cerebellar contribution to the disease, broadening our understanding of the complex neural involvements in MS (Mantel et al., 2022). In summary, the pathogenesis of Meige's syndrome appears to be associated with both a disproportionate ratio of different neurotransmitters in the brain and notable abnormalities in brain network functionality. However, the confirmation of these mechanisms is awaiting further research. Notably, previous studies have primarily concentrated on examining specific brain regions in MS patients. This focus has led to a gap in the literature, particularly regarding the analysis of centrality measures of each node within these brain networks. Moreover, while these studies have explored functional connectivity, there is a potential oversight in addressing higher-level functional connections at the network level. It is important to recognize that independent brain regions do not operate in isolation but rather form parts of interconnected networks, collaborating to perform common tasks.
In recent years, centrality analysis based on resting state fMRI has gained substantial attention and application in functional neurological research (Gao et al., 2021; Luo et al., 2021). This method includes voxel-wise centrality analysis, where each voxel is regarded as a node within the entire brain network. This graph-theoretical approach identifies certain brain regions that are more or less central within a global network, acknowledging the interconnected nature of human brain regions that collaborate to execute distinct cognitive functions (Bassett & Gazzaniga, 2011). During rest, these functional networks exhibit synchronized spontaneous activity, known as resting-state functional network connectivity (FNC) (Greicius et al., 2003).
In this study, we utilized degree centrality (DC) and eigenvector centrality (EC) as key features. Both DC and EC, grounded in graph theory, assess different aspects of brain network centrality. EC evaluates the relative importance of individual nodes within the entire brain network(Binnewijzend et al., 2014; Lohmann et al., 2010), while DC quantifies the connectedness of each node at a similar level(Di Martino et al., 2013). DC offers a direct method to describe the influence and function of nodes(Wu et al., 2020). Unlike DC, EC incorporates the centrality of adjacent nodes, and is relatively insensitive to motion artifacts, thereby minimizing the impact of slight head movements by MS patients on the results. Nevertheless, participants with head movements exceeding 2.0° were excluded during preprocessing (Binnewijzend et al., 2014; Lohmann et al., 2010; Wink et al., 2012). Independent component analysis (ICA) is particularly suited for analyzing resting-state fMRI due to its independence from specific temporal models or task-based designs. ICA can reveal brain functional features otherwise obscured due to the absence of prior information (McKeown et al., 1998). Spatial ICA, which produces maximally spatially independent, allows for the investigation of temporal dependencies within corresponding time courses. This property of Spatial ICA is especially beneficial for examining functional network connectivity (FNC), as it enables the analysis of weaker dependencies among Spatial ICA time courses, thus providing insights into FNC(Jafri et al., 2008). The centrality measures of individual nodes, a gap in previous research, are addressed through DC and EC methods in our study. Additionally, the exploration of higher-level functional connections at the network level is facilitated by FNC analysis using spatial ICA.
In this study, we employed centrality analysis, FNC, and mediation analysis of brain networks using resting state fMRI to address the aforementioned research questions. Our objectives were twofold: (1) to assess and characterize abnormalities in brain networks between MS patients and healthy controls (HC), with a focus on identifying changes in FNC and evaluating the significance and relative importance of nodes within the global brain network, and (2) to explore potential correlations between neuroimaging findings and clinical scales. To substantiate our hypotheses, we enrolled 28 HC and 31 patients diagnosed with MS, ensuring matching in terms of age, sex, and education level. Clinical parameters, including the disease course and the Burke-Fahn-Marsden Dystonia Rating Scale Movement scale (BFMDRS), were documented for all participants(Burke et al., 1985). Additionally, to discern whether the observed changes in brain function were attributable to dystonic origins or excessive facial movements, we included a comparative group of 28 patients diagnosed with hemifacial spasms (HFS), matched for age, sex, and educational background with the MS cohort.