Neuroimaging Study On Brain Functional And Structural Network Topological Properties Affected By Obesity And Bariatric Surgery

In recent years,with the rapid development of the economy,the development of internet catering and foreign fast food industries,people’s living standards and pace has also undergone significant changes.Long-term of unhealthy diets and poor exercise habits are bound to casue obesity.Obesity has been defined as a disease by the World Heath Organization（WHO）,and the complications related to obesity seriously endanger people’s physical and mental health.According to statistics,the number of obese people has surpassed 6 billion wordwide.In 2014,China’s obesity rank surpassed that of the United States and was ranked 1^st.Therefore,obesity has become one of the key issues we face.Clinically,bariatric surgery is the only long-term effective way to treat obesity.Neuroimaging studies have found that obesity causes significant changes in brain function and structure.Bariatric surgery can reduce the weight of obese patients,and also regulate abnormal brain function and promote neuroplasticity.Previous reports were mainly concerned with the changes in the brain regions of obese patients,and lacked overall research from the perspective of the brain network.In particular,changes in topological properties of obesity’s brain functional and structural network were reported less and the changes in topological properties of brain functional and structural network before and after bariatric surgery interventions have not been studied.Therefore,in the current study,obese patients were used as the research subjects,with bariatric surgery as the intervention,combined with graph theory and magnetic resonance imaging technology to study changes in topological organization of brain functional and structural network in them and after bariatic sugery.1)During resting state,the study of brain functional network found that the obese and normal weight groups both exhitied typical“small-world”properties.Comapred with normal weight group,the obese group showed significantly increased shortest path length（Lp）and decreased global efficiency（Eglob）;decreased nodal degree or efficiency was found in frontal（media orbitofrontal cortex-mOFC,rostral anterior cortex-rACC）,striatal（nucleus accumbens-NAc）,limic regions（insula,amygdala-AMY,hippocampus-HIPP,parahippocampus-PHIPP）and thalamus.Network-based statistics showed a sub-network with the rACC as the core node and weaker connections,composed of 31 nodes and 30edges.Longitudinal comparison of obese patients before and after surgery showed that the two groups also had typical“small-world”properties.The shortest path length and the normalized shortest path length were significantly decreased and the gloab efficiency was significant increased in one month after surgery.The nodal degree and efficiency were also increased in frontal（ventromedia prefrontal cortex-VMPFC）,cingulate gyrus（posterior cingutate cortex-PCC,anterior cingulate cortex-ACC）,limic regions（PHIPP）and so on.Meanwhile,some network attributes（such as Lp and Eglob）in the obese group were related to BMI.2)Based on the study of white matter network with diffusion tensor imaging technique,resuts showed that obese group and normal weight group both had typical“small-world”properties.Comapred with the normal weight group,increasd clustering coefficients（Cp）was found in the obese group;the nodal degree or efficiency of the prefrontal lobe was increased,including orbital part of the superior frontal gyrus（ORBsub）,orbital part of middle frontal gyrus（ORBmid）,and medial superior frontal gyrus（SFGmed）.In addition,the nodal degree of caudate was also increased.The nodal degree or efficiency of PHIPP,PCC and paracentral（PCL）were decreased.A sub-network with abnormal structural connections was defined based on network-based statistics in the obese group,including 34nodes and 31 edges,and the main nodes were located in the default mode network,limbic systerm,reward circuit,perceptual motion circuit and control network.Moreover,the NBS analysis of the longitudinal contrast before and after the surgery found that a sub-network with increased structural connections after surgery,composed of 15 nodes and 12 edges,and the involved brian regions were mainly in the default mode network,limbic system,reward circuit,control circuit,and supplemenyrary motor area.3)Selecting the differences about functional and structural network attributes between obese patients and normal subjects as the feature vector,we used a two-class classification method of support vector machine to classify functional and structural image data respectively,the results showed that the classification accuracy rate was 85.21%（sensitivity:88.38%;specificity:84.50%）and 81.26%（sensitivity:78.37%,specificity:81.80%）.This attempt to extract features from the network topological properties for classification will provide a priori knowledge and a theoretical basis for follow-up of weight reduction surgery outcome prediction and surgical screening for obese patients.The study demonstrates a abnormality of topological properties of the functional and structural network in obese patients,and the adjustment of topological properties of the functional and structural network by bariatric sugrey form the perpective of the global network,proving new neuroimaging evidence for understanding the pathogenesis of obesity and neural intervention mechanisms for bariatric sugrey,highlighting the potential for using network metrics as a disease-related imaging biomarker.