Advisory Committee Chair
Abidin Yildirim
Advisory Committee Members
Arie Nakhmani
Franklin Amthor
Leon Jololian
Document Type
Thesis
Date of Award
2019
Degree Name by School
Master of Science in Electrical Engineering (MSEE) School of Engineering
Abstract
The detection of neural spikes in real-time and accurately has become the center of interest for the researchers in the field of brain-machine/computer interface (BMI/BCI). The primary challenge in the Brain-Machine interface is to translate raw neuronal response signals into the control of electrical actuators. Only accurate and rapid classification of neural response can help efficiently and conveniently to disable peoples, particularly those suffering from spinal cord injury, stocks, etc., Recording from neurons and analyzing them with many different methods are not new. However, the primary challenge here is the real-time recording and classifying the spikes with higher accuracy. In particular, the last, i.e., accurate detecting and classifying the neural response is very paramount requirements for an efficient and functional BMI/BCI. In this approach, the slope values are found for each sample for creating spikes/non-spike templates. These templates are formed with a specific slope sequence and are fed into a fully connected Neural Network for classification. To evaluate this algorithm, two approaches were tested which are shape-based spike extraction and threshold-based spike extraction. Both approaches outperform in terms of efficiency and precision. Moreover, it also proves the methods for extracting a spike template without any known features of the spike.
Recommended Citation
Patel, Sahaj Anilbhai, "The Real-Time Extractiion Of Neural Spikes For Brain-Machine Interface Application Using Deep Learning Algorithm" (2019). All ETDs from UAB. 2680.
https://digitalcommons.library.uab.edu/etd-collection/2680
