You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster D26, Thursday, November 9, 6:15 – 7:30 pm, Harborview and Loch Raven Ballrooms

Neural network of verbal, nonverbal and amodal semantic processing deficits in semantic dementia

Yan Chen1, Keliang Chen2, Junhua Ding1, Yumei Zhang3, Qing Yang2, Yingru Lv2, Qihao Guo2, Zaizhu Han1;1Beijing Normal University, 2Fudan University, 3Beijing Tiantan Hospital

Semantic memory consists of three interactive principal components: amodal, verbal-specific and nonverbal-specific semantic processing, which are underpinned by separable neural networks. The study of patients with semantic dementia (SD) has emerged as an important lesion model for studying human semantic memory. Although it is well known of the amodal and modality-specific deficits of SD patients, the exact manner in which gray matter (GM) regions and their structural and functional connections are related to the deficits are not fully elucidated. The aim of this study was to map the neural network supporting the three semantic components in 33 SD patients. We acquired T1-weighted images, diffusion-weighted images, resting-state functional resonance images (rs-fMRI) and behavior data on word and picture semantic processing tasks. We found that the gray matter volume (GMV) of the left inferior temporal gyrus and the superior/middle temporal gyrus was highly correlated with verbal semantic processing, GMV of the right anterior middle temporal gyrus was associated with nonverbal semantic processing, while the GMV of the left anterior fusiform gyrus predicted amodal semantic processing. All six node pairs were found to be directly connected by WM tracts and the integrity of these six WM tracts was decreased in SD patients. We further found the integrity of different WM tracts was related with distinct semantic component. Rs-fMRI revealed the left anterior fusiform was functionally connected with the other three verbal and nonverbal nodes in healthy subjects. However, in SD patients, only one node pair (anterior fusiform-left superior/middle temporal gyrus) remained marginally significantly connected. No correlations were found between the strength of functional connections with distinct semantic components. Furthermore, the effects of regions and tracts remained significant even when we controlled for a wide range of potential confounding variables including overall cognitive state, whole brain gray matter volume and non-semantic control tasks (oral repetition and number calculation tasks). These results therefore address an important topic concerning the semantic deficits in SD, with implication for the neural network underlying different components of semantic knowledge in the human brain.

Topic Area: Language Disorders

Back to Poster Schedule