107. improving motor imagery brain computer interfaces with user response to feedback

Department: Electrical & Computer Engineering
Faculty Advisor(s): Virginia De Sa

Primary Student
Name: Mahta Mousavi
Email: mmousavi@ucsd.edu
Phone: 858-822-5095
Grad Year: 2018

Abstract
Many people suffer from motor disabilities and some with brain stem strokes and late stage neurodegenerative diseases are completely locked in and unable to voluntarily control their muscles. This means that without an intervention, they are unable to communicate - unable to express their thoughts, desires, and even their comfort or medical needs. However, in most locked-in patients, the brain and the autonomous nervous system are able to generate signals that can be read directly and brain computer interfaces (BCI) are a potential solution. Electroencephalography (EEG)-based brain computer interfaces (BCI) are high-speed, non-invasive, inexpensive and portable interventions that enable real-time control of a computer or robotic limb by analyzing electrical signals at the scalp. These brain signals however are combined with many sources of electrical noise and are also modified by changes in the user's emotions and thoughts that are not directly related to the task. In this study, we investigate the types of brain activity that interfere with the reading of EEG signals and develop algorithms to make use of these influences in favor of the BCI instead of letting the BCI to be negatively affected by them. The end goal is to have a much more robust and user-friendly communication system for paralyzed patients.

Industry Application Area(s)
Life Sciences/Medical Devices & Instruments | Software, Analytics

« Back to Posters or Search Results