Wearable Nanostructured Capacitive Sensor Systems for Multidimensional Intracranial Pressure, Sleep, and Fatigue Monitoring

Abstract

Sleep plays a central role in maintaining neurological, cardiovascular, and metabolic health, yet current sleep assessment methods remain limited in their ability to measure the brain’s physiological restoration. Conventional polysomnography is the clinical gold standard but is cumbersome and poorly suited for long-term or home-based monitoring. Home sleep apnea tests and consumer wearables improve accessibility but rely primarily on indirect cardiovascular or actigraphy signals, providing limited information about brain-specific physiology. Recent discoveries of the glymphatic system reveal that the brain’s waste-clearance pathways depend heavily on cerebrospinal fluid (CSF) transport, which is modulated by arterial pulsatility, vasomotion, respiration, and sleep stage. These physiological drivers are closely linked to intracranial pressure (ICP) dynamics. Prior studies have shown that ICP varies across NREM and REM sleep, responds to obstructive sleep apnea events, and reflects cerebrovascular flow. However, clinical ICP monitoring remains invasive and restricted to the intensive care unit. Existing noninvasive methods based on ultrasound, optical measurements, acoustic reflections, or skull deformation are not well suited for continuous, wearable, and comfortable sleep monitoring. To address this gap, this dissertation explores ICP as a promising biomarker of sleep quality and develops a wearable capacitive sensing system for continuous, noninvasive monitoring of regional ICP (r-ICP) using a single-electrode capacitance (SEC) sensing approach. Chapter 3 introduces an SEC sensing approach for noninvasive ICP monitoring and validates its feasibility using a porcine model. The work details the fabrication of carbon-nanotube composite paper (CPC) sensors, an SEC model that leverages the vascular network as a floating ground, and an experimental framework for measuring capacitance changes (ΔC) associated with r-ICP dynamics. The porcine model enables quantitative comparison of respiratory- and heartbeat-driven ΔC signals against simultaneously measured invasive ICP signals. Chapter 4 develops a wearable SEC sensing platform to monitor r-ICP during human laboratory tasks and posture transitions, thereby linking SEC-based measurements to physiological states relevant to ICP and sleep. The study characterizes posture-induced r-ICP changes during sit-to-supine transitions, head rotation, and Valsalva maneuvers. The significance of correlations between ΔC and electroencephalogram (EEG) signals is evaluated. Multidimensional ΔC patterns are examined to determine whether human SEC measurements reproduce key intracranial signatures identified in the porcine model. Chapter 5 introduces a sleep mask for r-ICP monitoring and evaluates its feasibility in a clinical study testing the hypothesis that r-ICP acts as a physiological sleep marker. The sleep mask provides comfortable, user-friendly monitoring of SEC signals and eye movements during nap time. ΔC behavior is characterized during sleep events, including apnea, REM transitions, and head rotation, and its relationship with EEG activity is analyzed. The study further examines associations between r-ICP features, subjective nap quality, and chronic sleep deficiency, and evaluates r-ICP-derived digital markers for classifying chronic sleep deficiency using statistical and machine-learning analyses. Chapters 6 and 7 investigate a wearable capacitive photic blink reflex (PBR) testing system for noncontact measurement of light-evoked ocular responses and its clinical application to chronic fatigue (CF) assessment. The system integrates CPC sensors and programmable LEDs into a glasses platform. In three healthy subjects, ocular capacitance (OC) signals are validated against electrooculogram (EOG) measurements, capturing dominant PBR responses with recovery times comparable to EOG and suggesting sensitivity to lateral differences in ocular response dynamics. The system is then evaluated in a clinical study of 27 participants, including 13 CF subjects and 14 healthy controls. OC analysis reveals reduced interocular synchronization in CF subjects during PBR responses, and OC-derived digital markers are developed to classify the CF group. These findings suggest that PBR-derived OC features can provide objective physiological markers for CF. In summary, this dissertation demonstrates a highly sensitive CPC sensor and a wearable sleep mask for sleep monitoring. The sensing principles of SEC-based r-ICP monitoring are investigated through sensor characterization, animal validation, and human studies. The clinical sleep study further explored r-ICP as a potential biomarker of sleep quality and chronic sleep deficiency. In addition, the fatigue study demonstrates a capacitive PBR monitoring device, potentially offering a novel diagnostic tool for chronic fatigue assessment.