Year: 2026
Type: dissertation
Abstract: This dissertation develops two practical systems toward real-world deployment: a braincomputer interface (BCI) for attention enhancement and a brain death determination (BDD) system for clinical decision support.For the attention-oriented BCI, (1) we propose a system architecture that extracts and identifies neural components related to attention and commands a drone to ascend/descend according to changes in attention level;(2) to efficiently extract and classify the relevant neural activity, we design computational algorithms for noise removal from recorded EEG, feature extraction of alpha and beta rhythms, and support vector machine (SVM) based classification; using data from 20 participants for training and a separate cohort of 4 for evaluation, we achieved a classification accuracy of 85.4%, a response time of 1.36 s, and a control success rate of 91.7%; (3) to realize the system, we designed a graphical user interface (GUI) with feedback capabilities and implemented a deployable BCI.For the BDD system, (1) we propose a dynamic computational algorithm based on EEG energy and EEG complexity, consistent with criteria used in brain death assessment; applying the developed algorithm to data from 64 pseudobrain-death cases including brain death and coma, the brain death determination accuracy reached 100%; (2) to support clinicians at the bedside, we designed an easy-to-operate GUI and implemented a realtime brain death determination system; additionally, we incorporated a large language model (LLM) to present patient symptoms and the supporting rationale to clinicians.Furthermore, we applied the proposed brain death determination algorithm not only to coma and brain death but also to anesthesia, confirming that analysis of EEG energy and EEG complexity across these three states enables objective quantification of the level of consciousness.
Authors: Gong, Jingming, キョウ, ケイメイ
Institution: Saitama Institute of Technology
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7162923531