| Objective:Based on multi-modal magnetic resonance imaging,to explore the similarities and differences of brain spontaneous brain activity and different types of networks in patients with cerebral small vessel disease(CSVD)with different olfactory function,and to analyze their correlation with olfactory and cognitive scores,so as to explore the neuroimaging mechanism of olfactory dysfunction in patients with CSVD,and to explore the potential clinical value of olfactory function assessment in CSVD.Methods:This study included 80 CSVD patients and 43 healthy control(HC)with similar age and sex.The basic data and the scores of Mini-Mental State Examination(MMSE),Montreal Cognitive Assessment(Mo CA),Chinese Academy of Sciences Psychological Research olfactory Test scale including odor threshold test(OTT),odor discrimination test(ODT)and odor identification test(OIT)of all subjects were collected,and clinical epidemiological analysis was carried out.According to OIT,patients with CSVD were divided into CSVD with normal olfactory function group(normal)and CSVD with dysosmia group(impaired).The patients were scanned with 3.0T magnetic resonance(GE Discovery MR75w)and 32 channel head coil.After preprocessing all the multimodal MRI data,the related indexes were calculated.1.Resting-state functional magnetic resonance imaging(rs-f MRI)imaging study of CSVD: based on the MA TLAB2013 b platform,use the options provided by DPARSF in the DPABI packages to calculate ALFF,f ALFF,Re Ho,and then perform Fisher-Z conversion,and then analyze the correlation with olfactory and cognitive scores.2.Different types of network analysis in CSVD:(1)construction of functional and structural networks;(2)Graph theory analysis and calculation of network indicators:Gretna is used to analyze the matrix,and calculate the relevant indicators of small world network,including average clustering coefficient(CP),path length(LP),Lp ratio of real network to random network(λ),CP ratio of real network to random network(γ)and ratio of γ to λ(σ).Global efficiency(Eglob)and local efficiency(Eloc)are calculated at the whole brain level.Nodal efficiency(Nodal E),degree centrality(DC),betweenness centrality(BC)are calculated at the node level.Results:1.rs-f MRI imaging study of CSVD: there were significant differences of ALFF in the right anterior cuneiform lobe between the three groups.In patients with CSVD,the ALFF in the right anterior cuneiform lobe was positively correlated with OIT(r=0.312,p=0.006),independent of sex,age,and years of education.There were significant differences of f ALFF in the left precentral gyrus and postcentral gyrus,but there was no correlation between f ALFF and clinical datas.There were significant differences of Re Ho in the left cuneiform lobe and the right dorsolateral superior frontal gyrus.In CSVD patients with normal olfactory function,Re Ho in the right dorsolateral superior frontal gyrus was negatively correlated with ODT(r=-0.386,p=0.032)and TDI(r=-0.380,p=0.035)independently of sex,age and years of education.2.Different types of network analysis of CSVD: in the functional network,the network indexes between groups was not obviously abnormal.In the structural network,the Eglob of CSVD patients was significantly lower than that of HC,and the correlation between Eglob and olfactory scores was influence with sex,age and years of education.There were abnormalities of Nodal E and DC in multiple brain regions in the structural network of CSVD patients.The node indexes of CSVD patients with different olfactory function were different absolutely,and they were significantly correlated with olfactory or cognitive scores.In CSVD patients with normal olfactory function,independent of age,sex,and years of education,Nodal E in left precentral gyrus was negatively correlated with OTT(r=-0.398,p=0.024),while Nodal E in right thalamus was positively correlated with OIT(r=0.377,p=0.033).Nodal E(r=-0.402,p=0.023)and DC(r=-0.432,p= 0.016)in left superior parietal gyrus were independent of age,sex,and years of education were significant negative correlation with MMSE scores.DC in the right anterior cuneiform lobe was positively correlated with OIT(r=0.374,p=0.035),independent of age,sex and years of education.In CSVD patients with dysosmia,Nodal E in the left postcentral gyrus was positively correlated with MMSE(r=0.430,p=0.004),independent of age,sex,and years of education.There was a significant positive correlation between Nodal E in the right superior parietal gyrus(r=0.335,p=0.030)and inferior parietal angular gyrus(r=0.501,p=0.001)and Mo CA,independent of age,sex and years of education.Nodal E in the left anterior cuneiform lobe was positively correlated with MMSE(r=0.378,p=0.014)and Mo CA(r=0.481,p=0.001)independently of age,sex and years of education.DC in the left anterior cuneiform lobe was positively correlated with Mo CA(r=0.367,p=0.017),independent of age,sex and years of education.Conclusions:This study showes that there are abnormalities in ALFF/f ALFF/Re Ho and Nodal E and DC of structural network in multiple brain regions in CSVD patients,and the abnormalities are associated with clinical data differently in CSVD patients with different olfactory functions.In CSVD patients without olfactory dysfunction,the result that Re Ho in the right dorsolateral superior frontal gyrus,Nodal E in the right thalamus and DC in the left anterior cuneiform of the structural network are related to the olfactory scores,suggests that the olfactory function of CSVD patients is not only directly involved in the olfactory related brain region thalamus,but also involved other brain regions such as the anterior cuneiform lobe and the dorsolateral superior frontal gyrus for olfactory processing.In CSVD patients with olfactory dysfunction,the result that Nodal E and DC in multiple nodes of the structural network are abnormal and only correlated with cognitive scores,and olfactory scores are significantly positively correlated with cognitive scores in this group of CSVD patients,indicates that the neuroimaging mechanism of olfactory dysfunction in CSVD patients is regulated by cognitive function. |
PDF Full Text Request