Current projects
Rule switching behaviour across species and computational modeling
Our model animals (marmosets, but also rats and mice) learn to perform tasks that require flexible decision making, motivated by a tasty treat of their choice. The tasks are the same or equivalent to those administered to human participants. Neural activities are recorded through wireless devices, which allows the animals to move freely; and their behaviours are monitored and quantified with state-of-the-art algorithms. This approach allows us to study the neural basis of cognitive flexibility, and to establish the causal link between neural activity and behaviour through circuit manipulation techniques.
In vivo electrophysiology
How does the brain represent the value of a choice, update it based on both positive and negative feedback and use it to guide actions? To answer this question, we will establish a mathematical model for reinforcement learning (RL) behaviour in mice, identify neural correlates for key parameters in the model, and manipulate the neural correlates to establish its causal link with behaviour. The results are then used to improve our models and generate new testable hypotheses.
The job of a neuron is to process information, but it does not do it alone. Using advanced data analysis techniques, we examine the dynamics at the level of neuronal ensembles, the interaction between different brain regions, and how these processes contribute to flexible decision making. We do this using a combination of newly acquired and existing data.
Population analysis of neuronal activities
Archived projects
Effects of ketamine on frontoparietal dynamics
N-Methyl-D-aspartate glutamate receptors (NMDARs) play a key role in prefrontal dynamics that support cognitive control. We simultaneously recorded local field potentials and single unit activities from lateral prefrontal (lPFC) and posterior parietal cortices (PPC) in macaque monkeys performing a rule-based antisaccade task. Not only altering the E/I balance in both areas, ketamine affected rule-coding in true oscillatory activities. It also reduced frontoparietal coherence and bidirectional connectivity.
Interestingly, a greater reduction in lPFC-PPC connectivity preceded a greater delay in saccadic onset under the rule-memorized, and a greater deficit in performance under the rule-visualized condition. We demonstrated the utility of acute NMDA receptor antagonist in understanding the mechanisms through which frontoparietal dynamics support cognitive control.