White Matter Structure Network Analysis In Patients With Type2 Diabetic Retinopathy

Objective:To investigate the white matter topological network changes（global properties,node properties and sub-networks）in patients with diabetic retinopathy（DR）,analyze the correlation between network changes and cognitive function score,and explore the structure network’s damage model and the possible influence on cognitive function,providing a structural basis for elucidating the neural mechanism of cognitive impairment in patients with DR.Methods:27 DR in-patients from the Department of Ophthalmology of the affiliated Hospital of Zunyi Medical University and 28 healthy volunteers matched in age,sex and years of education were recruited between October 2019 to October 2020.All the subjects had no mental or nervous system diseases and other eye diseases.MRI data were obtained with a 3.0T MRI scanner including T₂-FLAIR,3D-T₁WI,and DTI.Meanwhile,the clinical data,Mini-mental state examination（MMSE）and Montreal Cognitive Assessment（Mo CA）scores were collected.All the image preprocessing were performed with the PANDA.Network indexes（clustering coefficient,characteristic path length,global efficiency,local efficiency,standardized clustering coefficient,standardized characteristic path length,small world attribute value,node degree,node efficiency,node local efficiency）were calculated by GRETNA.The differences between groups were analyzed by SPSS.Pearson correlation analyses were used to demonstrate the relationships between network parameters in abnormal brain regions and cognitive function scores.For node index,P<0.011（1/90）was considered statistically significant,while for other data,P<0.05 was considered statistically significant.Results:（1）There were no significant differences in age,sex and years of education between DR and control group（P>0.05）,but the scores of MMSE（P=0.002）and Mo CA（P<0.001）were lower in DR group than control group.（2）Both groups of brain structural networks have small-world property.Compared with control group,the DR patients showed significantly increased standardized clustering coefficient（P=0.007）,characteristic path length（P=0.022）and small-world property values（P=0.019）,and decreased global efficiency（P=0.018）.（3）the DR patients showed reduced nodal degree in the left supplementary motor area,left cuneus,right lingual gyrus and right middle occipital gyrus,reduced node efficiency in the left calcarine fissure and surrounding cortex,left cuneus,left lingual gyrus,left superior occipital gyrus,bilateral middle occipital gyrus,right precuneus,left Thalamus,left heschl gyrus and left middle temporal gyrus,and reduced node local efficiency in the right posterior cingulate gyrus.（4）In DR patients,the alterd neural sub-networks are mainly located in the default mode network and visual network.（5）The nodal efficiency of left CAL is positively correlated with MMSE score（r=0.410,P=0.034）and Mo CA score（r=0.396,P=0.041）.Conclusion:This study reveals disrupted white matter network in DR patients,and correlation between nodal efficiency of left CAL and cognitive function scores,which provides a structural basis for cognitive impairment of DR and contributes to further understand the neural mechanism of cognitive dysfunction in DR.