Takuya Ito

This study introduces a neuroimaging framework to discover state-dependent signal flow hierarchy in the human cerebral cortex using integrated effective connectivity and signal flow mapping.

The paper investigates how frontoparietal and corticothalamic circuits contribute to relational reasoning by analyzing EEG data and neural field models during problem-solving tasks.

This study compares transformer models and human brain activity during relational reasoning tasks, finding alignment between attention mechanisms and neural networks involved in reasoning.

The paper investigates how brain activity patterns in alpha and beta rhythms within higher-order cortical networks track relational complexity during reasoning tasks.

The paper shows that network geometry influences how task representations change across different cortical regions, with low-dimensional connectivity compressing and high-dimensional connectivity expanding representations.

The paper demonstrates that distributed network flows can generate localized visual category selectivity in human visual cortex using fMRI data and computational models.

The paper examines how noise correlations in large-scale brain networks relate to signal correlations and their impact on information coding in a multi-task fMRI dataset.

The paper introduces causally informed activity flow models to better understand network-generated cognitive activations by using causal functional connectivity measures.

The paper estimates latent functional connectivity across multiple brain states and finds it better captures state-general intrinsic connectivity than resting-state measures.

The paper shows that global connectivity patterns in the brain predict the domain generality of cognitive control regions using resting-state fMRI data.

The paper shows that multi-task representations in human cortex follow a sensory-to-motor hierarchy with compression-then-expansion of dimensionality, supported by neural network models trained for sensory-to-motor transformation.

The paper quantifies the functional importance of task-state functional connectivity in shaping cognitive task activations and finds it improves prediction accuracy beyond resting-state connectivity.

The paper constructs neural network models from fMRI data to reveal how conjunction hubs support adaptive behavior through integrative computations.

The paper finds that whole-brain functional network organization is similar across mental disorders and healthy controls, but subtle differences predict diagnosis.

The paper shows that task states reduce neural correlations and variability across cortical areas in both primate and human data, increasing information coding capacity.

The paper investigates dynamic network properties of cognitive control networks in humans using a high-control-demand task, revealing distinct roles for the frontoparietal and cingulo-opercular networks.

The paper examines the relationship between localized task activity and distributed functional connectivity across the cortex, revealing a negative correlation and hierarchical differences in intrinsic timescales and myelin content.

This study shows that tACS at 10 Hz reduces BOLD adaptation in the human motion area and increases functional connectivity with the dorsal attention network.

The paper shows that task-evoked activations inflate task-state functional connectivity estimates and demonstrates that flexible modeling of hemodynamic responses effectively corrects this inflation.

The paper examines the relationship between depression symptoms and frontoparietal cognitive control network connectivity in healthy individuals, finding negative correlations with global connectivity.

The paper introduces information transfer mapping to show resting-state network topology predicts cognitive task information transfer between brain regions.

The paper identifies hierarchical activation states during rest and task conditions in fMRI data, revealing task-positive states occur frequently during rest.