Date of Award
Santa Clara : Santa Clara University, 2020.
The aim of this project was to find a way to differentiate active and rested brain signals in a patient using tasks without bodily movement to provide extremely motorly disabled patients a method of control for robotic devices that enable them to move independently of a caretaker. Although many control methods exist for less severely motorly impaired patients, this method would improve quality of life for all patients by allowing for movements to be controlled exclusively using the brain. The three steps for our project were to define the tasks and collect data, process the signals, and run the processed signals through a machine learning algorithm. In addition to the tasks not involving movement, having the subject’s eyes open was required as closing one’s eyes as a control method would not be practical. Different processing techniques were used to prepare the data and extract features for the training of the machine learning model for the classification task. Due to COVID-19, a limited amount of data was collected, resulting in inaccurate classification results. The “imagining-to-move” and “at rest” tasks that we designed for data collection appear to be the most effective when focusing on the mu rhythms at 7 to 12 Hz from the central cortex, but much more data is needed to prove this point. These tasks, brain area, and frequency ranges would be ideal for control method research projects in the future.
Baculi, Brent and Cansdale, Stuart, "Brainwave Classification for EEG-based Neurofeedback" (2020). Bioengineering Senior Theses. 91.