Investigation of the falling asleep dynamics and its closed-loop augmentation strategy
File(s)
Author(s)
Li, Junheng
Type
Thesis
Abstract
Sleep onset insomnia is a common sleep disturbance, particularly in the pathological ageing pop- ulation such as patients with neurodegenerative diseases and dementia. The bi-directional rela- tionship between sleep disorders and neurodegeneration makes it a novel early-stage treatment venue. Pharmacological treatments for insomnia often cause side effects and fail to show cog- nitive benefits. Novel neuromodulatory approaches, especially phase-synchronised closed-loop auditory stimulation targeting neural oscillations, have been promising candidates for enhancing deep sleep, but are yet to be explored for augmenting sleep onset.
Standard sleep staging has poor temporal resolution and bad matches to subjective experi- ence, especially for patients with sleep-onset insomnia. Therefore, a better understanding of the falling asleep dynamics outside the sleep staging is necessary for better outcome measurements and closed-loop control. This thesis aims to study the falling asleep brain dynamics measured via electroencephalogram (EEG) on large sleep datasets and test the feasibility of using closed-loop auditory stimulation to augment the process. I first developed a novel feature-based statisti- cal learning approach to explore the EEG dynamics and identified several whole-brain EEG biomarkers for distinguishing the wakefulness and sleep onset states. Combined with the dy- namical systems theory, I then built a functional model to describe the continuous falling asleep transition dynamics as a catastrophic bifurcation process. Additionally, I explored the regional brain dynamics and their interactions during the falling asleep process, which provided further insights into sleep initiation.
Lastly, I present a nap study to test the feasibility of phase-synchronised auditory stimulation in augmenting falling asleep, using the EEG biomarkers identified as outcome measurements. I also tested the effect of auditory-intervened naps on memory functions. The results showed a weak positive effect on associative memory. Further refined analyses are necessary to reveal mechanistic insights and guide future experimental explorations for a personalised stimulation design.
Standard sleep staging has poor temporal resolution and bad matches to subjective experi- ence, especially for patients with sleep-onset insomnia. Therefore, a better understanding of the falling asleep dynamics outside the sleep staging is necessary for better outcome measurements and closed-loop control. This thesis aims to study the falling asleep brain dynamics measured via electroencephalogram (EEG) on large sleep datasets and test the feasibility of using closed-loop auditory stimulation to augment the process. I first developed a novel feature-based statisti- cal learning approach to explore the EEG dynamics and identified several whole-brain EEG biomarkers for distinguishing the wakefulness and sleep onset states. Combined with the dy- namical systems theory, I then built a functional model to describe the continuous falling asleep transition dynamics as a catastrophic bifurcation process. Additionally, I explored the regional brain dynamics and their interactions during the falling asleep process, which provided further insights into sleep initiation.
Lastly, I present a nap study to test the feasibility of phase-synchronised auditory stimulation in augmenting falling asleep, using the EEG biomarkers identified as outcome measurements. I also tested the effect of auditory-intervened naps on memory functions. The results showed a weak positive effect on associative memory. Further refined analyses are necessary to reveal mechanistic insights and guide future experimental explorations for a personalised stimulation design.
Version
Open Access
Date Issued
2023-12
Date Awarded
2024-05
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Grossman, Nir
Barahona, Mauricio
Publisher Department
Department of Brain Sciences
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)