Search

Although semantic system is composed of two distinctive processes (i.e., semantic knowledge and semantic control), it remains unknown in which way these two processes dissociate from each other. Investigating the white matter neuroanatomy underlying these processes helps improve understanding of this question. To address this issue, we recruited brain-damaged patients with semantic dementia (SD) and semantic aphasia (SA), who had selective predominant deficits in semantic knowledge and semantic control, respectively. We built regression models to identify the white matter network associated with the semantic performance of each patient group. Semantic knowledge deficits in the SD patients were associated with damage to the left medial temporal network, while semantic control deficits in the SA patients were associated with damage to the other two networks (left frontal-temporal/occipital and frontal-subcortical networks). The further voxel-based analysis revealed additional semantic-relevant white matter tracts. These findings specify different processing principles of the components in semantic system.

Although the human temporal lobe has been documented to participate in semantic processing of both verbal and nonverbal stimuli, the exact neural basis underlying the common and unique processing of the two modalities is unclear. Semantic dementia (SD), a disease with a semantic-selective deficit due to predominant temporal lobe atrophy is an ideal lesion model to address this issue. However, many previous studies of SD used an impure patient sample or did not appropriately control for common components between tasks. To overcome these limitations, the present study aims to identify amodal semantic hubs and modality-specific regions in the temporal lobe by investigating behavioral performance on a verbal modality task (word associative matching) and a nonverbal modality task (picture associative matching) and neuroimaging data in 33 SD patients. We found that the left anterior fusiform gyrus was an amodal semantic hub whose gray matter volume correlated significantly with both modalities. We also observed two verbal modality-specific regions (the left posterior inferior temporal gyrus and the left middle superior temporal gyrus) and a nonverbal modality-specific region (the right lateral anterior middle temporal gyrus) whose gray matter volume correlated significantly with one modality when performance on the other modality was partialled out. The results remained significant when we excluded a wide range of potential confounding variables. Furthermore, to confirm the observed effects, we compared the performance of left- and right-hemispheric-predominant atrophic patients on the verbal and nonverbal tasks. The left-predominant patients showed more severe deficits in performance of the verbal task than the right-predominant patients, whereas the two groups of patients presented comparable deficits in the performance of the nonverbal task. These findings refined the structure of semantic network in the temporal lobe, deepening our understanding of the critical role of the temporal lobe in semantic processing.

BACKGROUND AND PURPOSE The diagnosis and monitoring of semantic variant primary progressive aphasia (sv-PPA) are clinically challenging. We aimed to establish a distinctive metabolic pattern in sv-PPA for diagnosis and severity evaluation. METHODS Fifteen sv-PPA patients and 15 controls were enrolled to identify sv-PPA-related pattern (sv-PPARP) by principal component analysis of 18 F-fluorodeoxyglucose positron emission tomography. Eighteen Alzheimer disease dementia (AD) and 14 behavioral variant frontotemporal dementia (bv-FTD) patients were enrolled to test the discriminatory power. Correspondingly, regional metabolic activities extracted from the voxelwise analysis were evaluated for the discriminatory power. RESULTS The sv-PPARP was characterized as decreased metabolic activity mainly in the bilateral temporal lobe (left predominance), middle orbitofrontal gyrus, left hippocampus/parahippocampus gyrus, fusiform gyrus, insula, inferior orbitofrontal gyrus, and striatum, with increased activity in the bilateral lingual gyrus, cuneus, calcarine gyrus, and right precentral and postcentral gyrus. The pattern expression had significant discriminatory power (area under the curve [AUC] = 0.98, sensitivity = 100%, specificity = 94.4%) in distinguishing sv-PPA from AD, and the asymmetry index offered complementary discriminatory power (AUC = 0.91, sensitivity = 86.7%, specificity = 92.9%) in distinguishing sv-PPA from bv-FTD. In sv-PPA patients, the pattern expression correlated with Boston Naming Test scores at baseline and showed significant increase in the subset of patients with follow-up. The voxelwise analysis showed similar topography, and the regional metabolic activities had equivalent or better discriminatory power and clinical correlations with Boston Naming Test scores. The ability to reflect disease progression in longitudinal follow-up seemed to be inferior to the pattern expression. CONCLUSIONS The sv-PPARP might serve as an objective biomarker for diagnosis and progression evaluation.

Background: Intrinsically organized large-scale brain networks and their interactions support complex cognitive function. Investigations suggest that the default network (DN) is the earliest disrupted network and that the frontoparietal control network (FPCN) and dorsal attention network (DAN) are subsequently impaired in Alzheimer's disease (AD). These large-scale networks comprise different subsystems (DN: medial temporal lobe (MTL), dorsomedial prefrontal cortex (DM) subsystems and a Core; FPCN: FPCNA and FPCNB). Our previous research has indicated that different DN subsystems are not equally damaged in AD. However, changes in the patterns of interactions among these large-scale network subsystems and the underlying cause of the alterations in AD remain unclear. We hypothesized that disrupted DN subsystems cause specific impairments in inter-system interactions and a higher regulatory burden for the FPCNA. Method: To test this hypothesis, Granger causality analysis (GCA) was performed to explore effective functional connectivity (FC) pattern of these networks. The regional information flow strength (IFS) was calculated and compared across groups to explore changes in the subsystems and their inter-system interactions and the relationship between them. To investigate specific inter-system changes, we summed the inter-system IFS and performed correlation analyses of the bidirectional inter-system IFS, which was compared across groups. Additionally, correlation analyses of dynamic effective FC patterns were performed to reveal alterations in the temporal co-evolution of sets of inter-subsystem interactions. Furthermore, we used partial correlation analysis to quantify the FPCN's regulatory effects. Finally, we applied a support vector machine (SVM) linear classifier to probe which network most effectively discriminated patients from controls. Results: Compared with controls, AD patients showed a decreased intra-DN regional IFS, which was significantly related to the inter-network's IFS. The IFS between the DN subsystems and FPCN subsystems/DAN decreased. Critically, the correlation values of the decreased bidirectional IFS between the DN subsystems and FPCNA diminished. Additionally, the Core and DM play pivotal roles in disordered temporal co-evolution. Furthermore, the FPCNA showed enhanced regulation of the Core. Finally, the MTL subsystem and Core were effective at discriminating patients from controls. Conclusion: The predominantly disrupted DN subsystems caused impaired inter-system interactions and created a higher regulatory burden for the FPCNA.

Background Semantic dementia (SD) is characterized by a selective decline in semantic processing. Although the neuropsychological pattern of this disease has been identified, its topological global alterations and symptom-relevant modules in the whole-brain anatomical network have not been fully elucidated. Objective This study aims to explore the topological alteration of anatomical network in SD and reveal the modules associated with semantic deficits in this disease. Methods We first constructed the whole-brain white-matter networks of 20 healthy controls and 19 patients with SD. Then, the network metrics of graph theory were compared between these two groups. Finally, we separated the network of SD patients into different modules and correlated the structural integrity of each module with the severity of the semantic deficits across patients. Results The network of the SD patients presented a significantly reduced global efficiency, indicating that the long-distance connections were damaged. The network was divided into the following four distinctive modules: the left temporal/occipital/parietal, frontal, right temporal/occipital, and frontal/parietal modules. The first two modules were associated with the semantic deficits of SD. Conclusion These findings illustrate the skeleton of the neuroanatomical network of SD patients and highlight the key role of the left temporal/occipital/parietal module and the left frontal module in semantic processing.

The ‘hub-and-spoke’ theory of semantic representation proposes that semantic knowledge is processed in a network comprising modality-specific regions connected to an amodal semantic hub. By studying semantic dementia, Chen et al. identify the semantic hub and its general and modality-specific white matter connections., The hub-and-spoke semantic representation theory posits that semantic knowledge is processed in a neural network, which contains an amodal hub, the sensorimotor modality-specific regions, and the connections between them. The exact neural basis of the hub, regions and connectivity remains unclear. Semantic dementia could be an ideal lesion model to construct the semantic network as this disease presents both amodal and modality-specific semantic processing (e.g. colour) deficits. The goal of the present study was to identify, using an unbiased data-driven approach, the semantic hub and its general and modality-specific semantic white matter connections by investigating the relationship between the lesion degree of the network and the severity of semantic deficits in 33 patients with semantic dementia. Data of diffusion-weighted imaging and behavioural performance in processing knowledge of general semantic and six sensorimotor modalities (i.e. object form, colour, motion, sound, manipulation and function) were collected from each subject. Specifically, to identify the semantic hub, we mapped the white matter nodal degree value (a graph theoretical index) of the 90 regions in the automated anatomical labelling atlas with the general semantic abilities of the patients. Of the regions, only the left fusiform gyrus was identified as the hub because its structural connectivity strength (i.e. nodal degree value) could significantly predict the general semantic processing of the patients. To identify the general and modality-specific semantic connections of the semantic hub, we separately correlated the white matter integrity values of each tract connected with the left fusiform gyrus, with the performance for general semantic processing and each of six semantic modality processing. The results showed that the hub region worked in concert with nine other regions in the semantic memory network for general semantic processing. Moreover, the connection between the hub and the left calcarine was associated with colour-specific semantic processing. The observed effects could not be accounted for by potential confounding variables (e.g. total grey matter volume, regional grey matter volume and performance on non-semantic control tasks). Our findings refine the neuroanatomical structure of the semantic network and underline the critical role of the left fusiform gyrus and its connectivity in the network.

Machine learning and pattern recognition have been widely investigated in order to look for the biomarkers of Alzheimer's disease (AD). However, most existing methods extract features by seed-based correlation, which not only requires prior information but also ignores the relationship between resting state functional magnetic resonance imaging (rs-fMRI) voxels. In this study, we proposed a deep learning classification framework with multivariate data-driven based feature extraction for automatic diagnosis of AD. Specifically, a three-level hierarchical partner matching independent components analysis (3LHPM-ICA) approach was proposed first in order to address the issues in spatial individual ICA, including the uncertainty of the numbers of components, the randomness of initial values, and the correspondence of ICs of multiple subjects, resulting in stable and reliable ICs which were applied as the intrinsic brain functional connectivity (FC) features. Second, Granger causality (GC) was utilized to infer directional interaction between the ICs that were identified by the 3LHPM-ICA method and extract the effective connectivity features. Finally, a deep learning classification framework was developed to distinguish AD from controls by fusing the functional and effective connectivities. A resting state fMRI dataset containing 34 AD patients and 34 normal controls (NCs) was applied to the multivariate deep learning platform, leading to a classification accuracy of 95.59%, with a sensitivity of 97.06% and a specificity of 94.12% with leave-one-out cross validation (LOOCV). The experimental results demonstrated that the measures of neural connectivities of ICA and GC followed by deep learning classification represented the most powerful methods of distinguishing AD clinical data from NCs, and these aberrant brain connectivities might serve as robust brain biomarkers for AD. This approach also allows for expansion of the methodology to classify other psychiatric disorders.

Lin Huang,1 Ke-Liang Chen,1 Bi-Ying Lin,1 Le Tang,1 Qian-Hua Zhao,1 Ying-Ru Lv,2,* Qi-Hao Guo1,* 1Department of Neurology, Huashan Hospital, Shanghai Medical College, Fudan University, Shanghai, China; 2Department of Radiology, Huashan Hospital, Fudan University, Shanghai, China *These authors contributed equally to this work Objectives: To find out whether the Chinese version of Montreal Cognitive Assessment Basic (MoCA-BC) and its subtests could be applied in discrimination among cognitively normal controls (NC), mild cognitive impairment (MCI), mild and moderate Alzheimer’s Disease (AD), and furthermore, to determine the optimal cutoffs most sensitive to distinguish between them.Design: A cross-sectional validation study.Setting: Huashan Hospital, Shanghai, China.Participants: There was a total of 1,969 participants: individuals with MCI (n=663), mild (n=345), moderate (n=441) AD, and cognitively NC (n=520) were recruited from the Memory Clinic, Huashan Hospital, Shanghai, China.Measurements: Baseline MoCA-BC scores were collected from firsthand data. Two subtests were calculated from MoCA-BC: the Memory Index Score of MoCA-BC (MoCA-BC-MIS) and the Non-memory Index Score of MoCA-BC (MoCA-BC-NM).Results: MoCA-BC was an effective cognitive tool to discriminate among NC, MCI, mild and moderate AD in the Chinese elderly across all education groups, implying that it was efficient not only for detecting MCI, but for different severities of AD as well. For MCI screening, the total score of MoCA-BC (MoCA-BC-T) and MoCA-BC-MIS had similar high sensitivity and specificity. For discrimination among MCI, mild and moderate AD, the MoCA-BC-T and MoCA-BC-NM had similar performance.Conclusion: MoCA-BC is an effective cognitive test to distinguish between NC, MCI, mild and moderate AD among the Chinese elderly with various levels of education. Keywords: mild cognitive impairment, Montreal Cognitive Assessment, Alzheimer’s disease, cutoff study

Individuals with semantic dementia (SD) typically suffer from selective semantic deficits due to degenerative brain atrophy. Although some brain regions have been found to be correlated with the semantic impairments of SD patients, it is unclear if the damage is actually responsible for SD patients’ semantic disorders because these findings were primarily obtained by examining the roles of local individual regions themselves without considering the influence of other regions that are functionally or structurally connected to the local individual regions. To resolve this problem, we investigated, from the brain network perspective, the relationship between the brain-network measures of regions and connections with semantic performance in 17 SD patients. We found that the severity of semantic deficits of SD patients was significantly correlated with the degree centrality values of the left anterior hippocampus (aHIP). Moreover, the semantic performance of the patients was also significantly correlated with the strength of gray matter functional connectivity of this region and two other regions: the left temporal pole/insula (TP/INS) and the left middle temporal gyrus. We further observed that the strength of the white matter structural connectivity of the left aHIP-left TP/INS tract could effectively predict the semantic performance of SD patients. When we controlled for a wide range of potential confounding factors (e.g., total gray matter volume), the above effects still held well. These findings revealed the critical brain network with the left aHIP as the center that could be contributing to the semantic impairments of SD.

Given that extensive cerebral regions are co-atrophic in semantic dementia (SD), it is not yet known which critical regions (SD-semantic-critical regions) are really responsible for the semantic deficits of SD. To identify the SD-semantic-critical regions, we explored the relationship between the degree of cerebral atrophy in the whole brain and the severity of semantic deficits in 19 individuals with SD. We found that the gray matter volumes of two regions [left fusiform gyrus (lFFG) and left parahippocampal gyrus (lPHG)] significantly correlated with the semantic scores of patients with SD. Importantly, the effects of the lFFG remained significant after controlling for the gray matter volumes of the lPHG. Moreover, the effects of the region could not be accounted for by the total gray matter volume, general cognitive ability, laterality of brain atrophy, or control task performance. We further observed that each atrophic portion of the lFFG along the anterior-posterior axis might dedicate to the loss of semantic functions in SD. These results reveal that the lFFG could be a critical region contributing to the semantic deficits of SD.

Delayed recall of words in a verbal learning test is a sensitive measure for the diagnosis of amnestic mild cognitive impairment (aMCI) and early Alzheimer's disease (AD). The relative validity of different retention intervals of delayed recall has not been well characterized. Using the Auditory Verbal Learning Test-Huashan version, we compared the differentiating value of short-term delayed recall (AVL-SR, that is, a 3- to 5-minute delay time) and long-term delayed recall (AVL-LR, that is, a 20-minute delay time) in distinguishing patients with aMCI (n = 897) and mild AD (n = 530) from the healthy elderly (n = 1215). In patients with aMCI, the correlation between AVL-SR and AVL-LR was very high (r = 0.94), and the difference between the two indicators was less than 0.5 points. There was no difference between AVL-SR and AVL-LR in the frequency of zero scores. In the receiver operating characteristic curves analysis, although the area under the curve (AUC) of AVL-SR and AVL-LR for diagnosing aMCI was significantly different, the cut-off scores of the two indicators were identical. In the subgroup of ages 80 to 89, the AUC of the two indicators showed no significant difference. Therefore, we concluded that AVL-SR could substitute for AVL-LR in identifying aMCI, especially for the oldest patients.