Smartwatch-like wearables have enabled seamless tracking of vital signs and physical activities, but still lack a significant feature: they are currently unable to provide any information about brain states or to modulate brain function for optimizing human health and performance. This project aims to make it possible for wearables to feature such capabilities. Being aware of brain states is not only extremely valuable in clinical studies but is also crucial to improving human performance in various everyday life activities. While recording neural signals directly from the scalp region is possible, it is impractical for use in everyday life. In order to fill this gap, the goal of this project is to pioneer a closed-loop brain-aware wearable architecture called MINDWATCH. This enables (1) decoding multidimensional brain states from noninvasive wearable devices and (2) applying corrective control. MINDWATCH will transform healthcare delivery (e.g., aging, autism, dementia) as well as human performance and productivity enhancement (e.g., online learning, smart workplaces). For instance, knowledge of mental health and cognitive engagement can enable detecting if a student is depressed or is not cognitively engaged/learning, which makes it possible to take corrective action early on. The research is integrated with educational and outreach activities with an emphasis on increasing the participation of minorities in science and engineering. These activities include hosting hands-on STEM K12 events, supervising undergraduate research interns and capstone senior design projects, creating educational videos, and interdisciplinary course development.
This project seeks to overcome the barriers to achieving brain-aware wearables by pioneering a transformative system-theoretic computational toolset for noninvasive closed-loop wearable architectures that monitor and modulate brain function without needing neural recordings. The proposed framework will (1) infer discrete brain-related events in real-world settings, (2) decode multidimensional latent neurobehavioral states based on inferred brain activity, and (3) apply robust adaptive control to maintain the neurobehavioral states within desired ranges. The closed-loop framework will be rigorously validated using experiments on interactive human-technology environments and mental health. The proposed research will provide foundational statistical signal processing and control-theoretic tools for the analysis of multimodal binary and continuous physiological observations that occur on multiple time-scales. While the initial focus is on two aspects of brain function, namely mental health and cognitive engagement, the proposed framework opens up the opportunity to investigate broader questions in computational neuroscience.
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.