EEG-Based Brain-Computer Interface for Transcription of Internally Generated Pitch, Poster 40
Abstract
The saddest pianist is one that cannot play. Unfortunately, motor neuron diseases can rob an individual of control of their body and produce dramatic reductions in quality of life. Brain-computer interfaces (BCI), however,... [ view full abstract ]
The saddest pianist is one that cannot play. Unfortunately, motor neuron diseases can rob an individual of control of their body and produce dramatic reductions in quality of life. Brain-computer interfaces (BCI), however, have the potential to dramatically increase patient quality of life by enabling an individual to communicate directly with a computer through self-directed alterations in their brain activity. In our project, we aim to identify characteristics of the electroencephalogram (EEG) that 1) may be subject to modulation by conscious thought and 2) can serve as potential signals for transcribing music directly from an individual’s brain activity. Initial explorations include recording EEG signals from a variety of cortical regions and using the statistical computing program Matlab to analyze power spectra to develop a framework for understanding putative signals that satisfy the conditions above. Once potential EEG targets are identified we intend to program Matlab to convert characteristics of these signals into specific pitches to enable the transcription of music directly from the individual’s brain activity. Ultimately, this approach could create the opportunity for disabled musicians to make music once again.
Authors
-
Harrison Hsiang '17
-
Michael Dash, Psychology
Topic Area
Science & Technology
Session
P2 » Poster Session 2 (2:45pm - Friday, 15th April, MBH Great Hall)