Exploring the use of wearables in the management of mild traumatic brain injury

Powell, Dylan (2022) Exploring the use of wearables in the management of mild traumatic brain injury. Doctoral thesis, Northumbria University.

Text (Doctoral thesis)
powell.dylan_phd(19014411).pdf - Submitted Version

Download (2MB) | Preview


a) Why is the subject of your thesis important?

Every year more than 1 million people attend Accident and Emergency with mild traumatic brain injuries (mTBI), many of which arise from Sports Related Concussion (SRC). Despite the high incidence of such injuries, there is still no gold standard method to monitor the wide variety of impairments (cognitive, visual, motor symptom) accompanying mTBI. Accordingly, there is concern about the long-term effects of mTBI if diagnosis is delayed or missed entirely. Current reliance on subjective techniques such as symptoms are non-specific and an unreliable indicator of recovery, making it difficult to know when it is safe for players to return to play (RTP). This highlights the need for testing and validating the accuracy and applicability of objective tools to aid diagnosis, monitoring, and RTP protocols for individuals exposed to mTBI and SRC.

b) How have you undertaken the research?

I have taken a systematic approach to this problem-based research, starting by understanding the clinical challenges of mTBI from SRC where amateur rugby union is used as an exemplar for investigation throughout the thesis. Both mTBI and SRC is an under-researched area confounded by insufficient medical staff available to recognise SRC and monitor players within low resource (community) based settings. This may place these individuals at an increased risk of having a delayed diagnosis or it being missed entirely. My hypothesis tests if the use of digital technologies may enable affordable mTBI management, ensuring continuity and objective personalised assessment to support traditional approaches. Accordingly, my thesis broadly comprises of a literature examination and preliminary validation and testing, progressing to an in-depth exploration involving larger datasets and concluding with recommendations for clinical practice.

c) What are your main research findings?

My multidisciplinary approach reveals that focusing on one impairment in mTBI is unlikely to reveal meaningful insight to mTBI/SRC and RTP. Instead, multimodal digital technologies could enable affordable management, ensuring consistency and continuity (e.g., between assessors) while offering objective personalised data to better support traditional approaches. My results provide insight and identify the usefulness of instrumented walking (gait) as a digital (bio) marker for mTBI management. Based on receiver operating characteristics (ROC) and area under the curve (AUC) analyisis free-living step velocity (i.e., walking speed) was the most sensitive (>0.72) at distinguishing healthy from acute SRC and may be useful for continuous monitoring and therefore informing SRC RTP.

In a purely computing science context, my findings have uncovered challenges and opportunities for further refinement. For example, there is still room for more ‘no code’ solutions in gait and algorithm analysis. Few clinicians would have the technical skillsets for completing free-living gait analysis. Therefore, validated algorithms within a "drag and drop", click and collect approach is needed to meet the recommend approach of remote, free-living monitoring of habitual behaviours. That is an important next step for the translation of academic research grade devices for broader deployment in clinical practice.

d) Why do your research findings matter

This thesis generally supports the suggested use of digital technologies as an affordable and objective method to support traditional approaches of assessment in mTBI/SRC. Passive and continuous monitoring solutions such as wearables are becoming ubiquitous in daily life. Moreover, the use of instrumented (lab) and free-living gait may fit that context with evidence of its use as a diagnostic tool. More work is needed to strengthen that claim as well as further investigate its use as a responsive tool. Identifying useful digital biomarkers in pathological cohorts such as mTBI may improve the detection of injuries and better inform safe (personalised) RTP guidelines. Identifying critical stages of recovery more accurately will also reduce the likelihood of premature return to play before full recovery, which is a necessary threshold in offering personalised care and rehabilitation. That is an important next step for the translation of academic research grade devices for broader deployment in clinical practice.

Item Type: Thesis (Doctoral)
Uncontrolled Keywords: sports related concussion, technology, digital biomarkers
Subjects: G400 Computer Science
Department: Faculties > Engineering and Environment > Computer and Information Sciences
University Services > Graduate School > Doctor of Philosophy
Depositing User: John Coen
Date Deposited: 29 Mar 2023 07:33
Last Modified: 29 Mar 2023 08:00
URI: https://nrl.northumbria.ac.uk/id/eprint/51552

Actions (login required)

View Item View Item


Downloads per month over past year

View more statistics