A Stroop Task FMRI Study In Patients With Vascular And Senile Cognitive Impairment

PurposeIn this study, used Bold-fMRI(blood oxygen level dependency - function magnetic resonance imaging) technology we observed and analyzed the cortex activation evoked by Stroop task in patients with SIVCI(subcortical ischemic vascular cognitive impairments), in order to inspect the influence on attention nerve circuit loop by ischemic impairment from small subcortical vessel lesion to enhance the understanding for the pathomechanism of subcortical ischemic vascular impairment and to find new indexes to reflect attention impairment in earlier period of vascular cognitive impairment. We studied the relationship and differences between brain activations of vascular, senile cognitive impairment patients and normal controls, in order to investigate the different pathomechanism between vascular and senile cognitive impairments and to appreciate the value of fMRI in the differential diagnosis of different cognitive impairments for improving their diagnostic quality.MethodsThe study was divided into two parts:1. The subjects included 10 patients with SIVCIND (subcortical ischemic vascular cognitive impairment no dementia), 10 patients with SIVD (subcortical ischemic vascular dementia) and 10 age, sex, and education-matched normal controls. The general cognitive functions of the subjects were assessed by Montreal cognitive assessment (MoCA) Beijing Version and mini-mental state examination (MMSE). Subjects performed the incongruent Stroop color-word task. In the test subjects were presented with a list of words"red","green", and"blue", however, the ink color of the words was discordant with the presented word. The control task was a solid white cross centered on the black background. The response time, incorrect response rate and no response rate were analyzed. All the patients and healthy volunteers underwent fMRI used the mode of stimulus-rest-stimulus when performing Stroop test. All images were taken with a Siemens Sonuta 1.5 Tesla MR scanner. Twenty T1-weighted axial slices (TE=13 msec, TR=500 msec, field of view=40×40 cm, 256×192 data matrix) were obtained parallel to the anterior–post commissure, which was identified with the aid of a sagittal localizer anatomical image. The functional image data were acquired with an echo-planar imaging sequence(64 x 64 matrix, 22 x 22cm field of view, echo time (TE) 40 ms, volume repetition time (TR)3000 ms, flip angle 90°). High-resolution structural imaging of the whole brain was performed with 3D gradient-echo, T1-weighted sequence, with the following parameters: 256x 256 matrix, inversion time (TI) 50.5 s, TE 56.3 ms, TR 511.7ms, flip angle 11°. Image processing and analyses were performed online with"T-test"and offline with Analysis of Functional Neuro Images (AFNI) software. The fMRI data were slice time-corrected. Echo planar images were coregistered to the image that minimized image translation and rotation relative to all other images. Corrected images were spatially filtered by using a Gaussian filter with a full width half maximum of 3 mm. The block design time-series was convolved to account for the ideal hemodynamic response function. For each voxel, a correlation coefficient was calculated, indicating the strength of relationship between the subjects'BOLD signal and the target reference function. The time-signal intensity curves and the functional images were obtained. After the functional and anatomical images for each participant were transformed into the Talairach and Tournoux coordinate system, the activation locations and areas of the SIVCIND, SIVD patients and normal control subjects were measured and analyzed.2. The subjects included 10 patients with mild cognitive impairment and 11 patients with Alzheimer's disease, who were age, sex, and education-matched with the subcortical ischemic vascular cognitive impairment no dementia and subcortical ischemic vascular dementia. The Stroop task performance and fMRI examination were the same as part 1. The response time, incorrect response rate and no response rate were analyzed. The activation areas and locations evoked by Stroop task were measured and analyzed. The relationship and differences between SIVCI, MCI and AD were analyzed.Results1. Subcortical ischemic vascular dementia patients had obvious cognitive disorder in visualspatial and executive function, attention, delayed memory, abstract, verbalization and orientation (P<0.05), and no cognitive disorder in naming compared with normal controls (P>0.05). Subcortical ischemic vascular cognitive impairment no dementia patients had obvious cognitive disorder in visualspatial and executive function, verbalization, attention and delayed memory (P<0.05), and no cognitive disorder in abstract, orientation and naming compared with normal controls (P>0.05).2. On Stroop color-word task SIVD patients showed higher incorrect response rate (P<0.05), higher omission rate (P<0.05) and longer reaction time (P<0.05) than those of normal control subjects. SIVCIND patients showed higher omission rate (P<0.05) and longer reaction time (P<0.05) than those of normal control subjects.3. For SIVCI groups bilateral anterior cingulate, DLPFC(dosolateral prefrontal cortex), VLPFC(ventralateral prefrontal cortex), inferior parietal lobe, occipital lobe and basal ganglia were activated. There was no difference of activation locations between SIVCIND and normal control subjects (P>0.05). The right basal ganglias for SIVD were less activated than those of normal controls (P<0.05). Compared to controls, SIVCIND patients showed distinctly increased prefrontal cortex activation, including bilateral DLPFC, VLPFC and right inferior parietal lobe (P<0.05), no difference in left inferior parietal lobe (P>0.05). SIVD patients exhibited decreased fMRI responses in bilateral DLPFC, VLPFC and inferior parietal lobe (P<0.05).4. There was no cortex activation difference between two hemispheres of normal control and SIVCIND subjects (P>0.05). However, for SIVD patients the activation areas of right inferior parietal lobe was less than that of the left (P<0.05). Pearson correlation\Spearman rank correlation analysis revealed that for SIVCI patients there were significant correlations between cortex activation areas of bilateral DLPFC, VLPFC, inferior parietal lobe and MoCA scores of total, attention, and visualspatial(P<0.05). There were significant correlations between cortex activation areas of bilateral DLPFC, left VLPFC and MoCA scores of language (P<0.05). There were significant correlations between cortex activation areas of bilateral DLPFC, inferior parietal lobe and MoCA scores of delay memory (P<0.05). There were significant correlations between cortex activation areas of right DLPFC, left VLPFC, right inferior parietal lobe and MoCA scores of orientation (P<0.05).5. On Stroop color-word task AD patients showed higher incorrect response rate (P<0.05), higher omission response rate (P<0.05) and longer reaction time (P<0.05) than those of normal control subjects. The MCI patients showed higher omission rate (P<0.05) and longer reaction time (P<0.05) than those of normal control subjects. However, there was no difference of incorrect response rate, omission rate and reaction time between MCI and SIVCIND, AD and SIVD patients (P>0.05).6. For AD and MCI groups bilateral anterior cingulate, DLPFC, VLPFC, inferior parietal lobe, occipital lobe and basal ganglia were activated. There was no difference of activation locations between MCI and normal control subjects (P>0.05). The right basal ganglias for AD were less activated than those of normal controls (P<0.01). Compared to controls, MCI patients showed distinctly increased cortex activation, including bilateral DLPFC and inferior parietal lobe (P<0.05), no difference in bilateral VLPFC (P>0.05). AD patients exhibited decreased fMRI responses in the regions of bilateral DLPFC and VLPFC (P<0.01), no difference in bilateral inferior parietal lobe (P>0.05). The activation areas of the two hemispheres for MCI and AD patients were not symmetrical. For MCI patients the activation areas of right inferior parietal lobe was less than that of the left (P<0.05). And for AD patients the activation areas of right VLPFC were less than that of the left (P<0.05).7. There was no difference of cortex activation locations evoked by Stroop task between MCI and SIVCIND, AD and SIVD patients (P>0.05). There was no difference of activation areas between MCI and SIVCIND patients (P>0.05). However, the activation areas of bilateral inferior parietal lobe for SIVD patients were less than that of the AD patients (P<0.05). So fMRI can not discriminate MCI and SIVCIND but can discriminate AD and SIVD by the activation areas of bilateral inferior parietal lobe.Conclusion1. SIVCI patients have obvious disorders in general assessment of cognitive function. Executive function, verbalization, attention and delayed memory are predominant in cognitive impairment of SIVCIND patients. SIVD patients have obvious cognitive impairment in executive function, attention, verbalization, delayed memory, abstract and orientation, but no cognitive disorder in naming.2. The results of computer-assisted Stroop test reveal that SIVCIND and MCI patients show higher omission rate and longer reaction time than those of normal control subjects. AD and SIVD patients show higher omission rate, incorrect response rate and longer reaction time than those of normal control subjects. However, the omission rate, incorrect response rate and reaction time can not discriminate SIVCIND and MCI, SIVD and AD.3. Our results suggest that for Stroop task the cortex activation areas include bilateral anterior cingulate, DLPFC, VLPFC, inferior parietal lobe, occipital lobe and basal ganglia. There is a bilateral DLPFC, VLPFC compensation in SIVCIND, and right basal ganglia, bilateral DLPFC, VLPFC and inferior parietal lobe dysfunction in SIVD, which suggests different neurophysiological characteristics between SIVCIND and SIVD. SIVCIND and SIVD have different fMRI characteristics and fMRI is a usefull tool in the examination and appreciation of the earlier period of vascular cognitive impairment.4. There was no cortex activation difference between two hemispheres of normal control and SIVCIND subjects. For SIVD patients the damage is not symmetrical and the left is the dominant hemisphere. There are significant correlations between cortex activation areas of bilateral DLPFC, VLPFC, inferior parietal lobe and MoCA scores of total, attention and visualspatial for SIVCI patients. The cortex activation of SIVCI during Stroop task performance in fMRI can reflect the cogniton of them well.5. There is a right basal ganglia, bilateral DLPFC and VLPFC cortex dysfunction in AD, and bilateral DLPFC and inferior parietal lobe compensation in MCI, which suggests different neurophysiological characteristics between AD and MCI. For MCI and AD the damage and compensation of the two hemispheres is not symmetrical and the left is the dominant hemisphere.6. There is no difference of cortex activation between MCI and SIVCIND patients. FMRI can not discriminate them. However, fMRI can discriminate AD and SIVD by the activation areas of bilateral inferior parietal lobe. The fMRI examination may have some potential value in the discrimination and assessment of different type of dementia.