Electronic Thesis and Dissertation Repository

Thesis Format



Master of Science




Owen, Adrian M.

2nd Supervisor

Debicki, Derek B.



There is a critical need for a bedside neuroimaging tool to aid in the prediction of functional recovery outcomes for patients with acute disorders of consciousness (DoC) in the early days following severe brain injury. Current neurobehavioral examinations and prognosis tools have limitations in predicting good outcomes, leading to potential mistreatment or premature withdrawal of life support. Functional near-infrared spectroscopy (fNIRS) is a viable candidate for such purposes due to its portability and cost-effectiveness. Auditory processing, viewed as a multi-level and multifaceted brain function, could provide a sensitive and specific marker of residual cognitive function in unresponsive patients. This study aimed to investigate the effectiveness of fNIRS for hierarchical assessment of auditory function and evaluate its applicability for predicting recovery outcomes in acute DoC. The capability of fNIRS for such an application was demonstrated by validating it against fMRI in a healthy population and cross-validating it in an entirely unresponsive patient with cognitive-motor dissociation. An innovative fNIRS-focused method was developed to quantify patients’ auditory function, and a data-driven method was explored to improve the sensitivity and specificity of auditory scores. Using these analytical tools, a direct association was found between auditory function and recovery outcome in a small patient cohort. Based on the study’s findings, the crucial role of methodological considerations in the use of fNIRS was discussed, and specific modifications in the stimulus and optical montage designs were suggested to enhance the method’s reliability.

Summary for Lay Audience

“What happens next?” This is the first question on everyone's mind when someone falls into a coma. Predicting the functional recovery outcome for comatose patients is a critical yet challenging question, even for critical care professionals. These predictions play an important role in decisions made for patients by medical professionals and loved ones, especially when deciding whether to withdraw life support.

Currently, established neurologic examinations and prognosis tools have some limitations, especially in determining if an adequate functional recovery is possible. Advanced neuroimaging methods can help assess the degree of preserved brain function in unresponsive patients.

Depending on its severity, brain injury may not damage the brain's ability to hear sounds but may damage regions involved in recognizing speech and extracting meaning from it. Therefore, we can determine the extent of damage by assessing brain activity in response to stimuli with different acoustic and linguistic features and using the assessment results to predict recovery outcomes. fNIRS is a portable method that can be used at the bedside for comatose patients whose transfer to imaging departments is risky. It uses red light to measure changes in blood oxygenation and estimate the activity in the brain's regions.

This research used fNIRS to record the brain responses of thirty healthy participants and eight unresponsive patients while listening to specific stimuli. The method was found to be suitable for detecting healthy participants’ brain responses. Using the reference activation maps from the healthy participants, we could detect preserved auditory function in an entirely unresponsive patient who suffered from an autoimmune condition that rendered her paralyzed but did not damage her brain function. Additionally, the degree of preserved auditory function in the patient cohort was found to be a predictor of their functional recovery outcome. We concluded that fNIRS has the capacity to be used as a tool aiding prognosis, for which further improvements in the method and its validation in larger patient cohorts are essential.