Static and dynamic interactions within the triple-network model in stroke patients with multidomain cognitive impairments

Yingying Wang,Hongxu Chen,Caihong Wang,Jingchun Liu,Peifang Miao,Ying Wei,Luobing Wu,Xin Wang,Peipei Wang,Yong Zhang,Jingliang Cheng,Siyuan Fan,Guifang Sun
DOI: https://doi.org/10.1016/j.nicl.2024.103655
2024-08-10
Abstract:Background: Internal capsule strokes often result in multidomain cognitive impairments across memory, attention, and executive function, typically due to disruptions in brain network connectivity. Our study examines these impairments by analyzing interactions within the triple-network model, focusing on both static and dynamic aspects. Methods: We collected resting-state fMRI data from 62 left (CI_L) and 56 right (CI_R) internal capsule stroke patients, along with 57 healthy controls (HC). Using independent component analysis to extract the default mode (DMN), executive control (ECN), and salience networks (SAN), we conducted static and dynamic functional network connectivity analyses (DFNC) to identify differences between stroke patients and controls. For DFNC, we used k-means clustering to focus on temporal properties and multilayer network analysis to examine integration and modularity Q, where integration represents dynamic interactions between networks, and modularity Q measures how well the network is divided into distinct modules. We then calculated the correlations between SFNC/DFNC properties with significant inter-group differences and cognitive scales. Results: Compared to HC, both CI_L and CI_R patients showed increased static FCs between SAN and DMN and decreased dynamic interactions between ECN and other networks. CI_R patients also had heightened static FCs between SAN and ECN and maintained a state with strongly positive FNCs across all networks in the triple-network model. Additionally, CI_R patients displayed decreased modularity Q. Conclusion: These findings highlight that stroke can result in the disruption of static and dynamic interactions in the triple network model, aiding our understanding of the neuropathological basis for multidomain cognitive deficits after internal capsule stroke.
What problem does this paper attempt to address?