| Summary: | <p>Understanding and measuring pain and brain development in neonates is essential to be able to provide the best care for this vulnerable population. This is particularly important for premature infants, for whom early life is filled with more painful procedures, and earlier exposure to extrauterine stimuli, which can adversely affect development. Infant pain assessments combine behavioural and physiological measures such as facial expression, crying, and heart rate. However, these metrics are not specific to pain experience, nor sensitive enough to provide reliable outcome measures for clinical trials to validate pain treatments in infants. Neuroimaging techniques provide means to study brain health, development and function. EEG and fMRI measurements of noxious-evoked brain activity could be used to develop more objective and specific pain assessment tools. This thesis focusses on using EEG and MRI to measure infant pain and its relation to overall brain development.</p>
<p>First, I present tests of the validity of an EEG template measure of noxious response in infants recruited at multiple hospital sites. EEG has been used to quantify noxious-evoked activity and study pain interventions in infants, but a standard generalisable approach needs to be established. I tested whether the EEG template discriminates between noxious and non-noxious stimuli, whether the scale of noxious response is equivalent across different hospital sites, and whether noxious response increases with age in premature infants. I found that noxious-evoked responses are significantly greater than non-noxious responses, but that the scale is not equivalent across study sites, and there was no significant age correlation. This suggests that the EEG template can be reliably used as a surrogate measure of pain, with promise for clinical trials. Additionally, data collection site should be accounted for as a confounding factor as needed.</p>
<p>Then, I focus on how MRI can aid our understanding of infant pain and the underlying neurophysiology behind differences in noxious-evoked activity. I present a machine learning model that I developed to predict the magnitude of noxious-evoked responses from resting- state brain activity in infants, using fMRI data. By applying this model to data from the independent Developing Human Connectome Project, I explore how predicted noxious- evoked responses relate to development metrics, including resting-state cortical function and microstructure, as well as prematurity, and assessments of infant cognitive and motor ability at 2-year follow up. I found that prematurity is associated with accelerated development of the nociceptive system, but disrupted neurodevelopment overall.</p>
<p>In summary, this thesis demonstrates the potential for neuroimaging techniques to improve our understanding of infant brain development, and improve clinical assessment and treatment of infant pain.</p>
|
|---|