Edge-centric functional network predicts risk propensity in economic decision-making: evidence from a resting-state fMRI study

Abstract Despite node-centric studies revealing an association between resting-state functional connectivity and individual risk propensity, the prediction of future risk decisions remains undetermined. Herein, we applied a recently emerging edge-centric method, the edge community similarity network...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Cerebral cortex (New York, N.Y. 1991) N.Y. 1991), 2023-07, Vol.33 (14), p.8904-8912
Hauptverfasser: Jiang, Lin, Yang, Qingqing, He, Runyang, Wang, Guangying, Yi, Chanlin, Si, Yajing, Yao, Dezhong, Xu, Peng, Yu, Liang, Li, Fali
Format: Artikel
Sprache:eng
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Abstract Despite node-centric studies revealing an association between resting-state functional connectivity and individual risk propensity, the prediction of future risk decisions remains undetermined. Herein, we applied a recently emerging edge-centric method, the edge community similarity network (ECSN), to alternatively describe the community structure of resting-state brain activity and to probe its contribution to predicting risk propensity during gambling. Results demonstrated that inter-individual variability of risk decisions correlates with the inter-subnetwork couplings spanning the visual network (VN) and default mode network (DMN), cingulo-opercular task control network, and sensory/somatomotor hand network (SSHN). Particularly, participants who have higher community similarity of these subnetworks during the resting state tend to choose riskier and higher yielding bets. And in contrast to low-risk propensity participants, those who behave high-risky show stronger couplings spanning the VN and SSHN/DMN. Eventually, based on the resting-state ECSN properties, the risk rate during the gambling task is effectively predicted by the multivariable linear regression model at the individual level. These findings provide new insights into the neural substrates of the inter-individual variability in risk propensity and new neuroimaging metrics to predict individual risk decisions in advance.
ISSN:1047-3211
1460-2199
DOI:10.1093/cercor/bhad169