Brain Computer Interfacing for Stroke Rehabilitation


Darren Leamy, Tomas Ward


It was long believed in the medical world that stroke caused irreversible damage to the brain. This would often mean that a sufferer may have lost the ability to move an arm, for example, and would never be able to move it again. Recent advancements in neurophysiology and related fields have refuted that idea, putting forth the idea that the brain, given the right conditions, can instead regain some of the lost function.lost function. This idea is called "Neuroplasticity".

The brain is partly made up of about 100 billion "neurons", each of them connected to potentially tens of thousands of other neurons. These neurons are responsible for everything we do and everything we think. Neurons grouped together in the brain usually work together at the same thing. However, most things we do (movement, sight, thought, memory) require different areas of the brain to work together. One area of the brain would send a signal to another area, which then sends a signal to another area etc. This means then that if one link in the chain is damaged, the function appears to have been lost. However, the function isn't completely lost, only a part of its process in the brain is. We now know that the brain can reorganize itself to compensate for the damaged area of the brain by strengthening weak inter-neuron connections and by forming new ones, thus regaining some of the lost function.

A Brain-Computer Interface (BCI) is a system for determining a person's intentions only by looking at signals generated in the brain. One common result of a stroke is the inability to control an arm. By monitoring the brain activity a BCI can tell if the stroke sufferer is trying to move their affected arm. The BCI can then direct an external device, such as a robotic arm, to move the arm for the patient. This completes a "bio-feedback loop" meaning that the brain is trying to move the arm and is also receiving signals telling it that the arm is indeed moving. We are investigating whether this biofeedback encourages neural plasticity. Given repeated attempts and repeated movement signals, the brain should rewire its connections to bypass the stroke-affected area of the brain and move the arm independently.

Papers (2011 - 2013)

Darren J. Leamy, Rónán Collins, and Tomas E. Ward. 2011. Combining fNIRS and EEG to improve motor cortex activity classification during an imagined movement-based task. In Proceedings of the 6th international conference on Foundations of augmented cognition: directing the future of adaptive systems (FAC'11), Dylan D. Schmorrow and Cali M. Fidopiastis (Eds.). Springer-Verlag, Berlin, Heidelberg, 177-185.

Leamy, Darren J., Tomás E. Ward, and Kevin T. Sweeney. "Functional near infrared spectroscopy (fNIRS) synthetic data generation." Engineering in Medicine and Biology Society, EMBC, 2011 Annual International Conference of the IEEE. IEEE, 2011.