Date of Award


Degree Type


Degree Name

Doctor of Philosophy in Electrical Engineering


Biomedical Engineering


Electrical, Computer, and Biomedical Engineering

First Advisor

Yalda Shahriari


A Brain computer interface (BCI) for communication bypasses standard efferent neural pathways towards peripheral nerves and muscles and directly translates brain activity into symbolic commands to facilitate communication in people with motor impairments. Although a range of access options have been developed, there are still many challenges, particularly for those with severe motor-control limitations. Limitations in availability of access options and the accuracy they provide may lead to patient frustration. Hence, many conventional non-invasive BCIs that rely on eye-gaze control and alternative communicative solutions, such as eye-tracking systems, fail to provide successful and efficient communication in people with severe motor deficit. In particular, despite great advances in state-of-the-art non-invasive BCIs, these systems are not sufficiently robust for practical long-term use due to inter-subject and session-to-session variability in the BCI performance. The motivation of this research is to incorporate cortical markers of cognitive disruptions to facilitate the development of reliable, robust, and person-centered BCIs for people with severe motor deficits. The research proposed in this work is centered on using functional near-infrared spectroscopy (fNIRS) recordings to compensate for the complexities which cause conventional visually triggered EEG-based BCIs, mainly since they fail when locked-in patients lose their oculomotor ability. Here, we have explored novel hemodynamic and electrophysiological cortical markers which, first, help to characterize cognitive disruptions and, then, to facilitate communication for people with ALS, as exemplar representatives of patients with severe motor deficits. Our obtained integrative signatures showed statistically significant between-group alterations in hemodynamic and electrophysiological makers evoked by a proposed visuo-mental (VM) dual-task, suggesting potential attentional, working memory, and decision-making dysfunctions, either in general task or task-specific demands in ALS patients. In our second work, our analyses demonstrated a shift toward a more centralized and asymmetric frontal network organization in ALS cohorts compared to controls. Furthermore, it was demonstrated that the global phase synchronization metric is associated with disease progression, and a few topological properties, including leaf fraction and tree hierarchy, are linked to disease duration. In the third work, we developed a new fNIRS-based BCI system in concert with a proposed VM paradigm to facilitate communication for ALS patients, particularly those in the later stages of the disease. Our findings indicated the potential efficacy of our proposed BCI system relying on fNIRS data for communication and control in ALS patients, particularly those in the later stages of their disease. In our final work, extending our previously developed fNIRS-EEG framework (chapter 2 to 4), we proposed a novel personalized experimental scheme in which a quick pre-screen was conducted prior to the main BCI protocol and then its extracted features were fed to previously constructed predictive platform based on training session data to select the subject and session specific variation of the task and then applying the appropriate correction to compensate for the systemic interfering factors. In sum, the proposed predictive scheme was fruitful in the subject-specific task/workload correction stage predicting correctly which of the variations of VM task leads to the highest accuracy for 81.82% of the subjects and the subject-adaptive correction strategies led to performance gains in the end.

Available for download on Thursday, November 17, 2022