| Project Area | Deciphering and Manipulating Brain Dynamics for Emergence of Behaviour Change in Multidimensional Biology |
|---|
| Project/Area Number |
22H05156
|
|---|
| Research Category |
Grant-in-Aid for Transformative Research Areas (A)
|
| Allocation Type | Single-year Grants |
|---|
| Review Section |
Transformative Research Areas, Section (III)
|
|---|
| Research Institution | Advanced Telecommunications Research Institute International |
|---|
Principal Investigator |
CORTESE Aurelio 株式会社国際電気通信基礎技術研究所, 脳情報通信総合研究所, 研究室長 (60842028)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
川人 光男 株式会社国際電気通信基礎技術研究所, 脳情報通信総合研究所, 所長 (10144445)
細谷 晴夫 株式会社国際電気通信基礎技術研究所, 脳情報通信総合研究所, 主任研究員 (50335296)
|
|---|
| Project Period (FY) |
2022-06-16 – 2027-03-31
|
| Project Status |
Granted (Fiscal Year 2025)
|
| Budget Amount *help |
¥108,420,000 (Direct Cost: ¥83,400,000、Indirect Cost: ¥25,020,000)
Fiscal Year 2026: ¥25,220,000 (Direct Cost: ¥19,400,000、Indirect Cost: ¥5,820,000)
Fiscal Year 2025: ¥24,960,000 (Direct Cost: ¥19,200,000、Indirect Cost: ¥5,760,000)
Fiscal Year 2024: ¥24,830,000 (Direct Cost: ¥19,100,000、Indirect Cost: ¥5,730,000)
Fiscal Year 2023: ¥14,170,000 (Direct Cost: ¥10,900,000、Indirect Cost: ¥3,270,000)
Fiscal Year 2022: ¥19,240,000 (Direct Cost: ¥14,800,000、Indirect Cost: ¥4,440,000)
|
|---|
| Keywords | adaptive behavior change / reinforcement learning / metacognition / neural dynamics / cerebellum / 適応的行動変容 / テンソル成分分析 / multimodal neuroimaging / Adaptive behavior change / decision-making / prefrontal cortex / neuroimaging / decoded neurofeedback / learning / probabilistic PCA / behavioral change / neurofeedback |
|---|
| Outline of Research at the Start |
Our research aims to revolutionise how we understand behaviour change. Over five years we will develop new computer algorithms that can very accurately analyse many behaviour measures from recordings of neural activity. Our team will apply machine learning techniques to complicated behavioural and neural data acquired from experiments in which mice and humans solve decision-making problems. We will then combine our new algorithms with advanced brain imaging in a novel experiment design, in both humans and mice, to control patterns of brain activity and cause changes in the targeted behaviour.
|
| Outline of Annual Research Achievements |
Research stream 1 (Hoang, Toyama, Kawato, Cortese): Our previous findings (Hoang et al. 2023) suggested that climbing fibre inputs encode reward-prediction errors, modulating Purkinje cell activity and refining motor behaviour. Building on this, our most recent study introduced three innovations. First, we applied Q-learning to model licking behaviour and extract reinforcement learning variables. Second, we performed regression analyses linking climbing fibre activity to reward and sensorimotor variables on a trial-by-trial basis. Third, we developed a cerebellar neural network model incorporating modular architecture with bidirectional plasticity. The results showed that distinct Purkinje cell modules could generate context-specific motor outputs based on reward-prediction error signals from climbing fibres (Hoang et al., 2025).
Research stream 2 (Okamoto, Six, Taylor, Ovadia, Oka, Gutierrez, Lobaskin, Cortese): We analysed and applied computational models - reinforcement learning and hierarchical Bayesian models - to study different confidence signals (perceptual, rule confidence) and their contributions to behaviour strategies (e.g. Okamoto et al. 2025). Importantly, we evaluated low-dimensional metacognitive representations in time and space across neuroimaging modalities (fMRI, MEG, ECoG). Beyond metacognitive signals, we also studied how beliefs about events transform and shape memories that later define behaviour changes (e.g., Cortese et al., 2024) We completed the neurofeedback experiment design.
We received the president award for our work related to neurofeedback.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
Research stream 1 (Hoang, Toyama, Kawato, Cortese): we constructed a large-scale spiking neuron network (5000 neurons) that includes various neuron types found within the cerebellum. This model has a modular structure to reflect the organizational principles observed in the cerebellar cortex. We investigated the role of positive feedback loops and of Purkinje cells, cerebellar nucleus neurons, and inferior olive neurons, in generating self-organizing network dynamics. Integration with a basal ganglia module enables to explore the interactions between these two key brain regions in learning from reward signals. In parallel, we began preparation with the Matsuzaki lab on the mouse DecNef project, defining the target cortical regions and neural population coverage, the decoding approach, and the behaviour dimensions to be modified through neurofeedback.
Research stream 2 (Okamoto, Six, Taylor, Ovadia, Oka, Gutierrez, Lobaskin, Cortese): Our current focus is characterising multidimensional confidence representations at behaviour, computational, and neural levels. To this end, we are analysing behaviour and neuroimaging data (fMRI, ECoG, MEG) from our novel unique decision task, in collaboration with the De Martino Lab at UCL, and developing a new hierarchical decision model based on multiple sources of noisy uncertainty. To extend our work on low-dimensional metacognitive neural signals, we are piloting a new task and neurofeedback protocol to be used in the next fiscal year. We are also moving ahead with organising the first international symposium on DecNef in July 2025.
|
| Strategy for Future Research Activity |
Research stream 1 (Hoang, Toyama, Kawato, Cortese): Our collaborators (Kitamura lab) are collecting two-photon calcium imaging data of climbing fibre inputs in mice performing an advanced Go/NoGo task. This new paradigm allows us to differentiate (1) motor errors, (2) cognitive errors, and (3) reward-prediction errors. We will apply and extend our previous framework to this new dataset by performing TCA to identify low-dimensional, task-relevant components of climbing fibre activity. Next, we will use trial-by-trial regression analyses to relate climbing fibre activity in each component to behavioural variables. Finally, we will refine our existing cerebellar network model by incorporating the new data and perform simulations. In parallel, we will initiate with the Matsuzaki lab the mouse DecNef project.
Research stream 2 (Okamoto, Six, Taylor, Ovadia, Oka, Lobaskin, Cortese): we will refine our computational models to our behaviour data to test the link between metacognition (confidence) and behaviour change measures. In particular, we will be focusing on how metacognition reflects multiple sources of error uncertainty. Our multimodal neuroimaging data (fMRI, MEG, ECoG) allows us to evaluate comprehensively low-dimensional metacognitive representations. We plan to collect new behaviour data with eye tracking to evaluate a new physiological correlate of confidence. In parallel, we will start piloting our new human decoded neurofeedback experiment for behaviour change. Finally, we will host the first international symposium on decoded neurofeedback in Japan in July 2025.
|
