Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

Alcohol’s Effect on Resting-State Brain Network Dynamics Using NASCAR Tensor Decomposition

Poster Session E - Monday, March 9, 2026, 2:30 – 4:30 pm PDT, Fairview/Kitsilano Ballroom

Kazuki Hachiken¹ (), Ayume Yamashita¹, Ryohei Nakayama¹, Kaoru Amano¹; ¹Graduate School of Information Science and Technology, The University of Tokyo

Alcohol consumption is known to influence cognition and brain activity. However, how alcohol consumption affects the organization and temporal fluctuation of brain network activity remains poorly understood. Previous studies have reported that certain brain networks are particularly susceptible to the effects of alcohol. However, the methods frequently used in those studies—such as independent component analysis (ICA) — cannot capture spatial overlaps or temporal interactions among networks, limiting understanding of how alcohol alters the competition or balance between brain networks. To overcome these limitations, we employed the Nadam-Accelerated SCAlable and Robust (NASCAR) tensor decomposition method, which simultaneously estimates spatial maps, temporal fluctuation in activity, and subject contributions. This approach allows the identification of spatiotemporally overlapping networks and clarification of alcohol effects on the interactions among networks. Using large-scale dataset from the Human Connectome Project (HCP), we first extracted robust interacting networks via NASCAR tensor decomposition, thereby forming a representative network atlas for a healthy population. In our experiment, 10 participants underwent two rs-fMRI sessions on separate days after consuming either alcohol or a soft drink (placebo).We then applied this network atlas to the acquired rs-fMRI data, estimating the subject contributions of each network. By comparing contributions score between the two conditions, we found decreased contribution of a network involving the default mode, attention, frontoparietal control, and salience systems in the alcohol condition compared with the soft drink (placebo) condition. These findings suggest that alcohol reduces the coordinated activity among networks related with attentional control and self-related processing.

Topic Area: OTHER