Dissertations and Theses

Date of Award

2024

Document Type

Dissertation

Department

Biomedical Engineering

First Advisor

Marom Bikson

Keywords

Vigilance, Attention, Sleep, EEG, Neuromodulation, Machine Learning

Abstract

The arousal spectrum is an integral part of human physiology, where oscillatory neural and physiological dynamics fluctuate between deep sleep and hyper alert states, over acute and diurnal cycles. Within the arousal spectrum, constructs like attention and vigilance aid in filtering and regulating ongoing exogenous sensory and endogenous neural activity, to maintain appropriate constant states of alertness.

Differing vigilant and attentional states can be assessed and characterized behaviorally with the use of verified questionnaires (KSS, PSQI) or psychophysical tasks (PVT, CTT); and physiologically with examinations of neural (EEG) and cardiographic (ECG/HRV) time and frequency dynamics. Failure of such mechanism can result in cognitive deficits, real-world consequences like accidents and injuries; and are often prominent in disorders like narcolepsy, insomnia, Alzheimer’s disease, autism spectrum disorders, age-related cognitive decline, and schizophrenia.

Noninvasive methods to alter vigilant/attentional states (at either end of the arousal spectrum) in human participants have been extensively explored including the use of electrical currents in techniques like transcranial electrical stimulation (tES). Evidence suggests that tES applied to the frontal cortices may interact with vigilant/attentional regulating mechanism and oscillations, where high frequency stimulation (>20 Hz) typically results in enhanced, whereas low frequency stimulation (<20 >Hz) typically results in diminished; attentional/vigilant outcomes. Such characterizations have lacked a well-rounded, multimodal examination of both discrete and continuous, and acute and long-term changes in vigilance and attention; behaviorally and physiologically.

This work examined the effects of High-Definition tES (HD-tES) on vigilance/attention at both ends of the arousal spectrum (toward increased and decreased vigilance/attention) with discrete and continuous metrics as well as multimodal behavioral (KSS, PVT, CTT, PSQI) and physiological (EEG, ECG, respiration) assessments.

The first aim of this work (Aim 1a) was to acquire, assess, and open source a multimodal dataset quantifying the effects of short-duration (30 sec), high frequency (30 Hz), frontal, HD-tES on acute (30 sec) changes in vigilance/attention during lowered vigilance/attentional states over continuous, extended periods of task performance (>30 mins). Results indicated that high frequency frontal HD-tES increased both behavioral (CTT) and neurophysiological (ECG) metrics of vigilance as compared to high frequency motor HD-tES; and all associated data have been open-sourced. As an exploratory aim (Aim 1b), we sought to assess the usability of acute physiological data (EEG, ECG) in predicting changes in behavioral measures of vigilance/attention with and without stimulation. Results indicated that physiological data can be used to predict stimulation responsiveness with the use of deep learning techniques. Our final aim (Aim 2), was to acquire, assess and quantify the effects of long-duration (10 mins), low frequency, frontal, HD-tES dose on prolonged (~10 mins) changes in vigilance/attention during an extended period of restful eyes closed (>10 mins). Results indicated that several doses of frontal HD-tES (namely 5 Hz and transcranial Endogenous Sleep Derived – tESD waveforms) were effective and non-disruptive in the transition to lowered vigilant states toward sleepiness as assessed by behavioral (PVT) and physiological (EEG, ECG/HRV) markers; however, some intervention effects were comparable to no stimulation.

Overall, this work indicates that HD-tES can be an effective intervention in altering attention and vigilance (enhancement and diminution).

Download

Available for download on Thursday, May 24, 2029

Share

COinS