No abstract

Objectives: We aim to investigate whether multi-dimensional diffusion tensor imaging (DTI) measures can sensitively identify different cognitive status of cerebral small vessel disease (CSVD) and to explore the underlying pattern of white matter disruption in CSVD.Methods: Two hundred and two participants were recruited, composed of 99 CSVD patients with mild cognitive impairment (VaMCI) and 60 with no cognitive impairment (NCI) and 43 healthy subjects as normal controls (NC). Full domain neuropsychological tests and diffusion-weighted imaging were performed on each subject. DTI metrics such as fractional anisotropy (FA), mean diffusivity (MD), the skeletonized mean diffusivity (PSMD), and structural brain network measures including network strength, global efficiency (EGlobal), and local efficiency (ELocal) were calculated. Region of interest (ROI) analysis of 42 white matter tracts was performed to examine the regional anatomical white matter disruption for each group.Results: Significant differences of multiple cognitive test scores across all cognitive domains especially processing and executive function existed among the three groups. DTI measures (FA, MD, and PSMD) showed significant group difference with the cognitive status changing. FA and EGlobal showed significant correlation with processing speed, executive function, and memory. ROI analysis found that white matter integrity impairment occurred from the preclinical stage of vascular cognitive impairment (VCI) due to CSVD. These lesions in the NCI group mainly involved some longitudinal fibers such as right superior longitudinal fasciculus (SLF-R), right superior fronto-occipital fasciculus (SFO-R), and right uncinate fasciculus (UNC-R), which might be more vulnerable to the cerebrovascular aging and disease process.Conclusions: DTI measures are sensitive neuroimaging markers in detecting the early cognitive impairment and able to differentiate the different cognitive status due to CSVD. Subtle changes of some vulnerable white matter tracts may be observed from the preclinical stage of VCI and have a local to general spreading pattern during the disease progression.

Deep learning methods have shown their great capability of extracting high-level features from image and have been used for effective medical imaging classification recently. However, training samples of medical images are restricted by the amount of patients as well as medical ethics issues, making it hard to train the neural networks. In this paper, we propose a novel end-to-end three-dimensional (3D) attention-based residual neural network (ResNet) architecture to classify different subtypes of subcortical vascular cognitive impairment (SVCI) with single-shot T2-weighted fluid-attenuated inversion recovery (FLAIR) sequence. Our aim is to develop a convolutional neural network to provide a convenient and effective way to assist doctors in the diagnosis and early treatment of the different subtypes of SVCI. The experiment data in this paper are collected from 242 patients from the Neurology Department of Renji Hospital, including 78 amnestic mild cognitive impairment (a-MCI), 70 nonamnestic MCI (na-MCI), and 94 no cognitive impairment (NCI). The accuracy of our proposed model has reached 98.6% on a training set and 97.3% on a validation set. The test accuracy on an untrained testing set reaches 93.8% with robustness. Our proposed method can provide a convenient and effective way to assist doctors in the diagnosis and early treatment.