Objective:To investigate the correlation between the location and volume of cerebral microbleeds (CMBs) and lacunar infarction (LI) in patients with cerebral small vessel disease (CSVD).Methods:Participants from the CSVD cohort in the Department of Neurology, Nanjing Drum Tower Hospital, the Affiliated Hospital of Medical School of Nanjing University from February 2017 to March 2019 were enrolled retrospectively. All participants underwent magnetic resonance imaging scans, standardized clinical assessment and diagnosis. AccuBrain, an automatic brain segmentation and quantification software developed by the Chinese University of Hong Kong, was used to quantitatively analyze the volume of CMBs and white matter hyperintensities (WMHs). Ordered multi-class logistic regression analysis was used to determine the independent influencing factors of LI, and then multiple linear regression analysis was used to investigate the correlation between the volume of deep or infratentorial CMBs (DI-CMBs) and the number of LI. Results:A total of 317 patients with CSVD were included in the analysis, including 214 (67.5%) in the non-LI group, 43 (13.6%) in the single LI group, and 60 (18.9%) in the multiple LI group. The comparison of the three groups showed that older age, male, smoking, drinking, history of previous stroke or transient ischemic attack (TIA), lower levels of high-density lipoprotein cholesterol, larger CMBs and WMHs volume, higher enlarged perivascular spaces (EPVS) grade might be the risk factors for LI. Ordinal multivariable logistic regression analysis showed that male (odds ratio [ OR] 2.058, 95% confidence interval [ CI] 1.084-3.909; P=0.027), previous stroke or TIA history ( OR 1.962, 95% CI 1.089-3.535; P=0.025), larger WMH volume ( OR 8.716, 95% CI 4.034-18.832; P<0.001), higher EPVS grade ( OR 1.915, 95% CI 1.292-2.839; P=0.001), larger DI-CMB volume ( OR 1.022, 95% CI 1.006-1.038; P=0.008) or more DI-CMB number ( OR 1.187, 95% CI 1.005-1.403; P=0.044) were the independent related factors of LI. Multiple linear regression analysis showed that there was a significant correlation between the volume of DI-CMB and the number of LI ( r=0.330, P<0.001). Conclusion:In patients with CSVD, there is a significant correlation between DI-CMBs and LI.

Objective To investigate the correlation between total MRI cerebral small vessel disease (SVD) score and overall cognitive functions and different cognitive domains.Methods Form January 2017 to June 2018,subjects aged 45-80 years and without dementia were recruited from the Neurology clinic,wards,Nanjing Drum Tower Hospital,the Affdiated Hospital of Nanjing University Medical School and communities.The overall cognitive function,executive function,processing speed,working memory,language function,visual-spatial ability,and depression were assessed with the related scale in all subjects.3.0 T MRI (T1-weighted imaging,T2-weighted imaging,diffusion-weighted imaging,liquid-attenuated inversion recovery sequence,and susceptibility-weighted imaging) was used to identify white matter hyperintensities,lacunar foci,microbleeds,and enlargement of perivascular space.The total SVD score was calculated.Results A total of 217 subjects were enrolled,including 24 normal middle-aged and elderly,65 at risk of cerebrovascular disease,128 with SVD changes in imaging findings.Multivariate linear regression analysis showed that after adjusting for possible confounding factors,the total SVD score and overall cognitive function (β =-0.105,95％ confidence interval [CI]-0.201--0.010;P =0.030),executive function (β =-0.135,95 ％ C!-0.216--0.054;P =0.001),and language function (β =-0.095,95％ CI-0.182--0.008;P =0.032) showed a significantly independent negative correlation.Conclusion The total SVD score was negatively correlated with the overall cognitive function,executive function,and impairment of language function.

Objective To explore the association between white matter hyperintensities volume and cognitive function in patients with hypertension.Methods Eighty-two patients with hypertension enrolled from the Department of Neurology of Nanjing Drum Tower Hospital from January 2017 to December 2017 were recruited in this study sequentially,who aged between 50 and 80 years.All subjects used automatic white matter quantitative technique to get their personal whole white matter hyperintensities volume,periventricular white matter hyperintensities volume and deep white matter hyperintensities volume.Neuropsychological scales were used to evaluate their global cognitive function,memory function,fluency,executive function and processing speed.The association between white matter hyperintensities volume and cognitive function via linear regression analysis was assessed.To explore whether the association between white matter hyperintensities volume and cognitive function be affected by the duration of hypertension,stratification analyses were conducted using the median of the duration of hypertension as separation.Results After adjusting for age,sex,education,lacunar infarctions and global cortical atrophy grade,the Trail Marking Test (2-1) score which reflects executive function was found to be related to whole white matter hyperintensities volume (standardized beta-value=0.268,P=0.019),periventricular white matter hyperintensities volume (standardized beta-value=0.243,P=0.037) and deep white matter hyperintensities volume (standardized beta-value=0.243,P=0.021).Victorial Snoop Test-2 score which reflects processing speed was found to be related to periventricular white matter hyperintensities volume (standardized beta-value=0.270,P=0.023).After additional adjusting for the duration of hypertension,the Trail Marking Test (2-1) score was still found to be related to whole white matter hyperintensities volume (standardized beta-value=0.277,P=0.016),periventricular white matter hyperintensities volume (standardized beta-value=0.251,P=0.032) and deep white matter hyperintensities volume (standardized beta-value=0.248,P=0.019).The Victorial Stroop Test-2 score was found to be related to whole white matter hyperintensities volume (standardized beta-value=0.237,P=0.041) and periventricular white matter hyperintensities volume (standardized beta-value=0.290,P=0.013).Conclusion Whole white matter hyperintensities volume,periventricular white matter hyperintensities volume and deep white matter hyperintensities volume are related to executive function,periventricular white matter hyperintensities volume is also related to processing speed in patients with hypertension,and the association between white matter hyperintensities volume and cognitive domain is not influenced by the duration of hypertension.

Objective To investigate the effect of hypertension on white matter microstructure and its correlation with cognitive function using automated fiber-tract quantification.Methods Consecutive subjects visited Drum Tower Hospital,the Affiliated Hospital of Nanjing University Medical School between January 2017 and July 2018 were collected.They were divided into hypertension without cognitive impairment (HTN-nonCI) group (n =44),hypertension with cognitive impairment (HTN-CI) group (n =50),and control group (n =25).The imaging data and neuropsychological scale test results of the subjects were collected.The automated fiber-tract quantification was used to obtain the diffusion parameters of 100 nodes on 20 fiber tracts in the whole brain.The differential segments of each fiber tract diffusion parameter between the control group and HTN-nonCI group,and the HTN-nonCI group and HTN-CI group were compared.Correlation analysis was performed in white matter fiber tracts with significant differences and each cognitive domain between the HTN-nonCI group and the HTN-CI group.Results The fractional anisotropy (FA) of white matter fiber tracts in 3 groups showed a decreasing trend,and the mean diffusion diffusivity (MD) showed an increasing trend.The comparison between the control group and the HTN-nonCI group showed that there were significant differences in the FA values of the midpoint of left thalamic radiation tract partial to brainstem,the genus of corpus callosum near brainstem,and frontal part and and proximal ventricle of splenium of corpus callosum (all P＜0.05);there were significant differences in the MD values of the middle part of left thalamic radiation tract and near the brainstem,the right thalamic radiation tract near the brainstem,the top of left corticospinal tract and near brainstem,the middle hippocampus of right cingulate tract,the middle part of genu of corpus callosum,the splenium of the corpus callosum near the lateral ventricle,the left uncinate tract near the forehead (all P ＜ 0.05).A comparison between the HTN-nonCI group and the HTN-CI group showed that there was significant difference in the FA value between the distributed segments of cingulate gyrus of left cingulate tract and the left inferior fronto-occipital tract near the occipital lobe,in which cingnlate gyrus of left cingulate tract was significantly correlated with the Montreal Cognitive Assessment Scale score (standardized3 =0.268,P =0.029);there were significant differences in the right thalamic radiation tract near the brainstem,the forehead and proximal ventricle of splenium of the corpus callosum,and the scattered distribution segments of the right inferior longitudinal fasciculus,in which the right inferior longitudinal fasciculus was significantly correlated with memory (standardizedβ=-0.243,P=0.047) and executive function (standardizedβ=-0.284,P=0.021).Conclusions Microstructural integrity of white matter was generally destroyed in patients with hypertension,but some segments were more susceptible to hypertension.The integrity of cingnlate gyrus of cingulate tract was significantly correlated with the overall cognitive function.The integrity of inferior iongitudinal fasciculus was significantly correlated with the executive function and memory.

Objective:To study the correlations of hippocampal subfield volumes with cerebrospinal fluid (CSF) biomarkers in Alzheimer's disease (AD) patients.Methods:Forty-nine patients with cognitive impairment, including 30 AD patients and 19 non-AD patients, were recruited in Department of Neurology, Drum Tower Clinical College, Nanjing University of Chinese Medicine from May 2017 to December 2021. Concentrations of Aβ 1-42, total tau protein and phosphorylated tau protein in CSF were analyzed by enzyme-linked immunosorbent assay (ELISA). Volumes of 12 hippocampal subfields were calculated using FreeSurfer image analysis. Differences of clinical data, neuropsychological scores and CSF biomarker concentrations between the 2 groups were compared. Partial correlation was performed to analyze the correlations of volumes of hippocampal subfields with CSF biomarker concentrations. Results:AD patients had significantly lower Mini-mental State Examination (MMSE) scores and Aβ 1-42 concentration in CSF than non-AD patients ( P<0.05); AD patients had significantly lower volumes of the right hippocampal parasitulum, dentate gyrus and CA4 than non-AD group ( P<0.05); the right parasubiculum volume was negatively correlated with CSF Aβ 1-42 ( r=-0.445, P=0.023) and positively correlated with CSF P-tau ( r=0.393, P=0.047) in AD patients. Volumes of left hippocampus tail, parasubiculum, CA1, molecular layer, dentate gyrus, CA3 and CA4 were negatively correlated with CSF total tau ( P<0.05). No significant correlation was noted between hippocampal subfield volumes and CSF biomarker concentrations in non-AD patients. Conclusion:Some right hippocampal subfields in AD patients atrophy compared with those in non-AD patients with cognitive impairment; the right parasubiculum may play a compensatory role in disease process, while volumes of the left hippocampus decreased with increased CSF total tau.

Objective:To explore the correlations of brain network functional connectivity (FC) alterations with cerebrospinal fluid (CSF) pathological biomarkers in patients with Alzheimer's disease (AD).Methods:A total of 39 patients with cognitive impairment, admitted to Department of Neurology, Drum Tower Hospital, Affiliated Hospital of Medical School, Nanjing University from January 2020 to December 2022 were recruited; 23 patients were with AD and 16 with non-AD. Clinical data were compared between the 2 groups. Resting-state functional MRI (rs-fMRI) data were collected, and FC differences between brain networks and FC differences within brain networks were compared by independent component analysis. Correlations of FC differences between brain networks and FC differences within brain networks with concentrations of β-amyloid protein 1-42 (Aβ 1-42) and Tau protein in CSF were analyzed. Results:Compared with the non-AD group, AD group had significantly lower Aβ 1-42 in CSF ( P<0.05). Compared with those in the non-AD group, FC alterations between the left frontoparietal network (lFPN) and anterior default mode network (aDMN) and between the visual network (VN) and posterior cingulate cortex (PCC), as well as FC alterations in lFPN, were significantly increased in AD group ( P<0.05). Compared with those in the non-AD group, FC alterations between lFPN and cerebellar network (CEN), and FC alterations in aDMN, sensorimotor network (SMN) and VN were significantly decreased in AD group ( P<0.05). In AD group, FC in SMN was positively correlated with total Tau and phosphorylated-Tau181 in CSF ( P<0.05); FC between VN and PCC was positively correlated with total Tau in CSF ( P<0.05). CSF Aβ 1-42 was positively correlated with FC alterations in aDMN and VN, but negatively correlated with FC in FPN ( P<0.05). Conclusion:In AD patients, characteristic changes in FC within and between multiple brain networks are noted, which are related to changes of Tau protein and Aβ 1-42 in CSF.

Objective　To examine the use of Montreal Cognitive Assessment (MoCA) scale for prediction of cognitive outcome.Methods　Data from 286 elderly patients who had performed cognitive assessment twice or more were obtained retrospectively,and divided into Normal Control (NC),Mild Cognitive Impairment (MCI) and dementia groups according to their baseline performance.The correlations between each item scores of MoCA and total score change at follow up were compared within groups.Results　Total score change was significantly correlated to the MMSE,cube,attention and subtraction in NC group;visuospatial/executive,clock-hands and digit-backward in MCI group;and abstraction in dementia group.Conclusion　Cognitive prognosis is influenced by the course of disease,interventions,learning effect and regression to the mean.Executive function decline can be considered as the early sign of Alzheimer’s disease.