Biology subjects

Onoda, K.

Publications and source records attributed to Onoda, K..

3 recordsLinked to original sources

bioRxiv2026

Collapse of local circuit integrated information {Phi} during NREM sleep

Clarifying the mechanisms underlying the emergence of consciousness remains a fundamental challenge in modern neuroscience. Integrated Information Theory (IIT) provides a mathematical framework derived from the phenomenological properties of consciousness as its axioms. IIT proposes that consciousness is identical to a systems intrinsic cause-effect information structure, quantified by integrated information {Phi}. While IIT predicts that the {Phi} of a neuronal system should decrease during the loss of consciousness, this hypothesis has remained untested at the neural circuit level. The present study provides empirical support for this IIT prediction. It was found that {Phi} within local circuits decreases during non-rapid eye movement (NREM) sleep compared to wakefulness and REM sleep, regardless of cortical laminar structure or brain regions, and even after controlling for firing rates. The reduction in {Phi} was particularly pronounced during OFF-periods, when neural activity is collectively suppressed. These results align with IITs framework, suggesting that the loss of consciousness corresponds to a collapse of integrated information within local circuits that cannot be explained solely by changes in individual system elements (such as overall firing rates). The findings provide empirical support for IIT as a mathematical theory aiming to explain conscious experiences.

neuroscience↗

bioRxiv2022

Expected Costs of Mental Efforts are Updated When People Exert Effort, not by Prospective Information

To understand why people avoid mental effort, it is crucial to reveal the mechanisms by which we learn and decide about mental effort costs. This study investigated whether mental effort cost learning aligns with temporal-difference (TD) learning or alternative mechanisms. Model-based fMRI analyses showed no correlation between cost prediction errors (CPEs) and activity in the dorsomedial frontal cortex/dorsal anterior cingulate cortex (dmFC/dACC) or striatum at the time of a fully informative effort cue about upcoming effort demands, contradicting the TD hypothesis. Instead, CPEs correlate with dmFC/dACC (positively) and caudate (negatively) activity at effort completion. Furthermore, only activity patterns at effort completion predict subsequent choices. These results show that decision policies are updated retrospectively at effort completion, updating expected costs with prediction error between experienced effort and prior expectations, demonstrating mental effort cost learning is retrospective, and imply that adaptive learning of mental effort cost does not follow canonical TD learning. Significance StatementUnderstanding how people learn about mental effort costs is essential for advancing theories of motivation and cognitive control. However, the algorithms supporting such learning remain unclear. This study addressed this gap and found that temporal-difference learning, commonly used to explain reward learning, could not account for how people learn about effort. Instead, decision policies were updated retrospectively at effort completion, based on a prediction error between experienced effort and prior expectations. These findings reveal that mental effort cost learning is fundamentally retrospective and imply that it relies on mechanisms distinct from canonical temporal-difference learning.

neuroscience↗

bioRxiv2022

Comparing Deep Learning Models for Age Prediction Based on the Resting State fMRI Dataset from the "Brain Dock" Service in Japan

In Japan, many hospitals provide the unequaled service of medical check-up called the "Brain Dock"; however, there is a paucity of studies aimed at leveraging functional magnetic resonance imaging (fMRI) in hospitals. We obtained the resting-state fMRI (rs-fMRI) scans of about 695 patients who accessed this service for preventive medicine against Alzheimers Disease. In this study, we created deep learning models for age prediction with the ultimate aim of arriving at standard protocols for introducing rs-fMRI into clinical settings, particularly in Brain Docks. With that view, an assemblage of modeling conditions was attempted, changing multiple parameter values, features based on data extraction methods (region of interest-wise mean blood-oxygen-level-dependent lump-sum time series models or dynamic functional connectivity models), deep learning algorithms (Transformer, Multi-task Transformer, and unidirectional or bidirectional long short-term memory models), and different atlas-dependent brain region segmentation methods (including the Automated Anatomical Labeling and Harvard-Oxford atlases). As a result, a robust and highly significant correlation was obtained between actual and predicted ages from all types of methodologies. In addition, we determined that some conditions had a relatively large impact on prediction performance based on extended comparisons. The accuracy decreased, particularly according to the choice of atlases but with the same modeling conditions. Notwithstanding, we found that atlases based on intrinsic functional connectivity provided significant prediction accuracy even with a small number of regions to a similar extent as networks lowered spatial granularity. Moreover, we found that multi-task learning with other phenotype data (related to gender differences) was possible, but did not improve the prediction accuracy as much as expected. Despite these limitations, our results could provide a hopeful prospect of introducing fMRI into the field of neuro-clinical practice. HighlightsO_LIAge prediction modeling was performed using the rs-fMRI data of a Japanese "Brain Dock" service targeting mainly elderly people, which made prediction more challenging. C_LIO_LIA robust and highly significant correlation was obtained from all types of methodologies changing data extraction methods, deep learning algorithms, and brain atlases. C_LIO_LIAge prediction was successful even with a very small number of regions exclusively based on intrinsic functional connectivity networks, although there was a difference in the achievement of modeling. C_LI

neuroscience↗