Description
- Abstract:
- Intracortical brain computer interfaces directly link the brain to external devices. When a person attempts to move, electrodes inserted into the motor cortex can record neural activity. These neural signals can be decoded to estimate the person's intended action. Through the technology, users with tetraplegia have controlled a computer cursor as well as a robotic arm. Here, several algorithms for better decoding are developed and evaluated. A total of four innovations are investigated. First, an approach based upon penalized maximum likelihood is shown to mitigate the negative effects of neural signal instabilities. Second, detecting when the decoding procedure makes an error is analyzed. Third, decoding outputs are mixed in a novel way to improve cursor control. Finally fourth, a nonparametric hierarchical decoder improves accuracy over current best practice. The performance gains are demonstrated using data variously from simulation, clinical research sessions with people with tetraplegia, and experiments with monkeys.
- Notes:
- Thesis (Ph.D. -- Brown University (2014)
Access Conditions
- Rights
- In Copyright
- Restrictions on Use
- Collection is open for research.
Citation
Homer, Mark L.,
"Novel Algorithms for Better Decoding of Neural Signals for Intracortical Brain Computer Interfaces"
(2014).
Biomedical Engineering Theses and Dissertations.
Brown Digital Repository. Brown University Library.
https://doi.org/10.7301/Z07M069Z
Relations
Collection:
-
Biomedical Engineering Theses and Dissertations
Theses and Dissertations for the Biomedical Engineering department....