Funded Projects

Understanding the relationship between structure and function in the human brain is a key interest in neuroscience. In recent years the focus is turning to understanding the role of the white matter in human cognition, brain function and neurological disorders.

Electrodes implanted in the brain have great potential, with applications in neurodegenerative disease, brain-computer interfaces, and more. However, the presence of electrodes in brain tissue causes a response known as gliosis, in which a scar forms around the electrode, reducing its effectiveness and access to neurons.

Stroke is the main cause of adult disability; 80% of survivors sustain motor (movement) deficits that interfere with activities of daily living. There exists no proven therapeutic strategy for motor recovery of the upper extremity following stroke.

A hallmark of the motor system is its ability to execute different skilled movements as the situation warrants, thanks to the flexibility of motor learning. Despite many behavioral studies on motor learning, the neural mechanisms of motor memory formation and modification remain unclear.

In the course of everyday functioning, animals (including humans) are constantly faced with real-world environments in which they are required to shift unpredictably between multiple, sometimes unfamiliar, tasks. But how brains support this rapid adaptation of decision making schema, and how they allocate resources towards learning novel tasks is largely unknown both neuroscientifically and algorithmically.

Yi Lui's project aims to use deep brain microstimulation (DBMS), which causes even less brain damage and has higher spatial resolution than DBS, for memory recovery.

Dr. Brandon Jay Bhasin will use engineering principles from modern control theory, experimental neuroscience and computational neuroscience to significantly advance understanding of how feedback driven plasticity in a tractable neural circuit is orchestrated across multiple synaptic sites and over various timescales so that circuit dynamics are changed to improve performance.

The project aims to alleviate this bottleneck by developing a weak supervision system that optimally deals with time-series data and takes advantage of multiple data modalities.

We propose a novel framework for efficient Bayesian cognition called Inference via Abstraction (IvA), which learns to approximate complex world models with simpler abstractions that capture main dependencies, but leverage structure in the prior distribution for efficient inference. We instantiate IvA with a combination of probabilistic graphical models and deep neural networks.