CAREER: Neural Dynamics of Sleep-Mediated Learning in Brain Computer Interface (BCI) Applications

Virginia Polytechnic Institute And State University, Blacksburg VA

Investigators

Abstract

Brain Computer Interfaces (BCIs) allow an individual’s brain activity to control an external device or application (e.g., a prosthetic arm, wheelchair, or cursor on a computer screen), but learning to use BCIs proficiently often takes significant time and effort, which can be frustrating to the user. A better understanding of the role of sleep in BCI learning may allow for interventions that speed up learning in BCI applications. Sleep is known to boost learning and memory consolidation for a variety of tasks and sleep can be manipulated to enhance and speed up learning (e.g., by playing task-related auditory cues during sleep). However, the importance of sleep for BCI learning remains unknown. This research will examine whether sleep plays a role in learning BCI tasks and whether sleep can be manipulated using covert auditory sounds to speed up BCI learning. This work will provide the basic science knowledge needed for a safe, non-invasive, sleep-based intervention that will allow individuals to master the use of an assistive BCI device more quickly by engaging learning pathways that are active during sleep. An enhanced understanding of the fundamental relationship between sleep and learning will also lead to better treatment for diseases in which cognitive and memory deficits are tied to abnormal sleep (e.g., schizophrenia). This project also offers research opportunities that will enhance undergraduate and K-12 education through curriculum development and outreach initiatives that incorporate aspects of sleep and learning, BCIs, and brain activity. This project will employ computational and experimental methods to investigate the role of sleep in BCI learning. Subjects will move a cursor by modulating their brain activity before and after sleep. Simultaneously, neural data will be collected: either EEG data from healthy subjects (to probe large-scale neural dynamics) or human intracranial depth electrode data from patients undergoing invasive monitoring (to probe circuit-level and single neuron dynamics). In additional studies using the same basic design, task-related or carefully chosen non-task-related auditory cues will be played covertly during sleep. These studies will identify the neural dynamics crucial to sleep-mediated learning of BCI applications and the ways in which to optimally engage these processes and manipulate sleep to speed up learning. The generalization of the adaptive stimulation paradigm developed for this work could be applied in broad contexts in which stimuli are presented to manipulate, probe, and characterize brain dynamics; to facilitate such investigations, the software will be made freely available. The proposed work will lay the foundation for a large-scale and potentially transformative research program to investigate sleep and learning, as the experimenter’s ability to choose and perturb the mappings that control BCI tasks may facilitate the identification of the neural features critical for sleep-mediated learning and potentially reveal novel mechanisms related to sleep-mediated learning. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.