Sixty-two AIS patients admitted to Weifang People's Hospital (Shangdong, China) between September 2014 and August 2015 were selected using computer-generated random number matching for this study. The final cohort consisted of 34 males and 28 females between 45 and 80 years (mean, 58.7±8.8 years). Inclusion criteria were as follows: i) Meeting the Chinese and WHO AIS diagnosis and treatment criteria, with MRI and CT confirmation; ii) onset ≤7 days before examination; and iii) able to complete research project and the MoCA. Exclusion criteria were as follows: i) Brain trauma, brain tumor, cerebral aneurysm, cerebrovascular developmental malformation, cerebral hemorrhage, and other craniocerebral diseases; ii) underlying diseases, such as heart, lung, kidney, or liver dysfunction; iii) severe condition predictive of <12-month survival; and iv) pre-existing cognitive impairment due to Alzheimer's disease or other neuropsychiatric diseases. This study was approved by the Weifang People's Hospital Ethics Committee, and informed consent was obtained from all the patients or their relatives.

Patients underwent brain CT scan with a scanning seam thickness of 5 mm and layer spacing of 5 mm. After injection of the contrast medium, CT scanning of the medial temporal lobe, bilateral frontopolar lobes, parietal lobe, and basal ganglia was performed.

Patients underwent 3.0-T MRI in the supine position with eyes closed. Patients wore earplugs to reduce scanning noise, and head motion was restricted using foam pads. Patients were requested to keep quiet during scans, including T1-weighted, T2-weighted, and FLAIR sequences, to assess WML severity/location and to exclude other encephalopathies. Resting-state functional MRI was conducted using the echo-planar imaging (EPI) sequence (seam thickness, 3 mm; axial view, 36 layers; voxel size, 3.5×3.5×3.0 mm³; matrix, 64×64; repetition time (TR), 2,000 msec; echo time (TE), 30 msec; flip angle (FA), 90°; field of view (FOV), 222×222 mm). High-resolution T1-weighted structural images were collected using a 3D MP-RAGE sequence (seam thickness, 1 mm; sagittal view, 176 layers; voxel size, 1.0×1.0×1.0 mm³; matrix, 256×256; TR, 1,900 msec; TE 2.52 msec; inversion time (TI), 900 msec; FA, 9°; FOV, 256×256 mm).

CT images were imported into a processing station. MRI data were processed by SPM8 running on the MATLAB R2012a platform. Post-processing of MR images included time correction, dynamic correction, high-resolution structural image segmentation, regression removal of whole-brain signals, space standardization, spatial smoothing, de-systematic linear drift correction, band-pass filtering, and standardized calculation of amplitude of low-frequency fluctuation (ALFF) value.

The MoCA evaluates 8 cognitive domains: Visuospatial/executive function, memory, naming, language, attention, delayed recall, abstract thinking, and orientation. The total score is 30 points and a score of 26 or lower is indicative of cognitive impairment.

Lesions were scored from 0 to 3 according to the following patterns: i) No reduced WM signals (0 points); ii) reduced WM signals in occipital angle and the edge of frontal angle of lateral ventricles (1 point); iii) reduced WM signals in occipital angle surroundings and the frontal angle of lateral ventricles with extension to the semioval center (2 points); and iv) reduced WM signals around the whole lateral ventricles and extending to the semioval center with blending (3 points). Lesion severity was scored according to the following changes in signal strength: i) No reduction in signal (0 point); ii) mildly reduced signals (1 point); iii) moderately reduced signals (2 points); and iv) substantially reduced signals (3 points).

Periventricular lesions were scored according to the following patterns: i) No lesions (0 points); ii) pencil-like or cap-like thin lesions (1 point); iii) smooth haloes at lesion site (2 points); and iv) irregular periventricular high signals extending to deep WM (3 points). Deep WMLs were scored according to the following patterns: i) No lesions (0 points); ii) punctate separate lesions (1 point); iii) fused lesions (2 points); and iv) large fused lesions (3 points). The total score was obtained by adding the individual periventricular and deep WML scales together.

Measurement zones within four WM regions were demarcated on axial T2-weighted images of the lateral ventricles and the third ventricle: i) The semioval center (4 separate zones: Bilateral anterior and posterior zones); ii) corona radiata (6 separate zones: Bilateral anterior, posterior, and cingulate gyrus zones); iii) high external capsule layer (6 separate zones: Bilateral anterior, posterior, and cingulate gyrus zones); and iv) low external capsule layer (4 separate zones: Bilateral anterior and posterior zones). The WML severity in each zone was scored as: i) Normal (0 points); ii) <50% of region involved (1 point); and iii) >50% of region involved (2 points). Severity scores were weighted based on cholinergic fiber distribution density. The weighting coefficients were 1 for the semioval center, 2 for the corona radiata, 3 for the high external capsule layer, and 4 for the low external capsule layer. The score for each region was obtained by adding all points for the measurement of small zones and multiplying by the weighting coefficient. The total score was calculated as the sum of scores for all four regions (maximum final score of 100, with 50 points/hemisphere).

SPSS 19.0 software was used for data processing. Measurement data with normal distributions are expressed as mean ± standard deviation (SD). Group means were compared by t-test. Enumeration data are expressed by case number (%). Proportions were compared by χ² test. The strengths of correlations were assessed by Spearman's correlation analysis. A P<0.05 was considered to indicate a statistically significant difference.

The mean Blennow scale score was 1.6±0.5 (maximum 3) for all 62 subjects, of which 33 were scored as 0 or 1 and 29 as 2 or 3 (indicative of more severe and extensive WMLs). MoCA score was significantly higher in the less severe 0–1 point Blennow scale group compared to the 2–3 points group, and the proportion of subjects scoring <26 points on the MoCA (defined as cognitive dysfunction) was significantly higher in the 2–3 points group (P=0.034) (Table I). The MoCA score was negatively correlated with Blennow scale score (r=−0.326, P=0.002).

The mean Fazekas scale score for periventricular plus deep WMLs was 3.4±0.8 (maximum 6) for the entire 62-patient cohort, of which 30 patients scored 0–3 and 32 scored 4–6 points. Again, MoCA score was significantly higher in the 0–3 points Fazekas scale group (with less severe and extensive WMLs), and the proportion of patients with cognitive dysfunction according to MoCA score <26 points was significantly higher in the 4–6 points Fazekas scale group with more severe periventricular/deep WMLs (P<0.01) (Table II). In addition, MoCA score was also negatively correlated with the Fazekas scale score (r=−0.404, P=0.031).

The mean CHIPS score (maximum 100) was 65.7±12.5 for all 62 subjects, of which 25 scored 0–50 points and 37 scored 51–100 points. MoCA score was significantly higher in the 0–50 points group than in the 51–100 points group, and the proportion with cognitive impairment (MoCA score <26 points) was significantly higher in the 51–100 points CHIPS group (P<0.05) (Table III). MoCA score was negatively correlated with CHIPS score (r=−0.234, P=0.042).