AUTHOR=Wei Na , Deng Yiming , Yao Li , Jia Weili , Wang Jinfang , Shi Qingli , Chen Hongyan , Pan Yuesong , Yan Hongyi , Zhang Yumei , Wang Yongjun TITLE=A Neuroimaging Marker Based on Diffusion Tensor Imaging and Cognitive Impairment Due to Cerebral White Matter Lesions JOURNAL=Frontiers in Neurology VOLUME=Volume 10 - 2019 YEAR=2019 URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2019.00081 DOI=10.3389/fneur.2019.00081 ISSN=1664-2295 ABSTRACT=Background: The peak width of skeletonized mean diffusivity (PSMD) is a new, fully automated, and robust imaging marker for cerebral small vessel disease (SVD). It is considered to be strongly associated with processing speed. However, it has not been applied to cerebral white matter lesions (WMLs) yet. Our study aimed to investigate the correlation between PSMD and cognition, particularly executive function, which has emerged as the most prominently affected cognitive domain in patients with WMLs. Methods: A total of 111 WML patients and 50 healthy controls (HCs) were enrolled, and their demographic information and cardiovascular disease risk factors were recorded. Subjects were divided into three groups: WMLs with normal cognition (WMLs-NC), WMLs with vascular cognitive impairment (WMLs-VCI), and HCs. They underwent conventional head MRI and DTI scans followed by neuropsychological and psychological examinations, including tests of Montreal Cognitive Assessment (MoCA) and executive function. We compared the difference in executive function and PSMD among the three groups and analyzed the correlation between PSMD and cognitive function in all subjects. Results: There were no significant differences in demographic characteristics (age, gender, level of education, and cardiovascular disease risk factors) among the three groups (P>0.05). However, there were significant differences in global cognition (P<0.0001), executive function (P<0.0001), and PSMD (P<0.0001) among the three groups. The averaged PSMD value (×10-4mm2/s) was 2.40±0.23, 2.68±0.30, and 4.51±0.39 in the HC, the WMLs-NC, and the WMLs-VCI groups, respectively. There was no correlation between PSMD and cognition in the HC group. PSMD was significantly correlated with MoCA scores (r=-0.3785, P<0.0001) and executive function (r=-0.4744,P<0.0001) in the WMLs-NC group and in the WMLs-VCI group (r=-0.4448, P<0.0001 and r=-0.6279, P<0.0001, respectively). Conclusions: WML patients have higher PSMD and worse cognitive performance than do healthy controls. PSMD is strongly associated with global cognition and executive functions in WML patients. This result provides new insights into the pathophysiology of cognitive impairment in WML patients. PSMD could be a surrogate marker for disease progression and can thus be used in therapeutic trials involving WML patients.