Quantitative MRI T2 Relaxometry and Physical Activity in Acute Mild Traumatic Brain Injury: An Exploration of the Measurement and Modification of Potential Neuroinflammation

Abstract

Mild traumatic brain injuries (mTBI) account for approximately 90% of all traumatic brain injuries (TBI) and are often associated with prolonged symptoms and functional impairments. Despite their prevalence and significant physical, psychosocial and financial impact, the underlying pathophysiology and factors influencing recovery remain poorly understood. mTBI typically results from a sudden impact, acceleration, deceleration, or rotational force, triggering a cascade of secondary injuries. A key component of this cascade is neuroinflammation, which, while initially protective, can lead to chronic deficits and neurodegeneration if prolonged or excessive. Despite the known role of inflammation in brain injury, significant gaps remain in understanding its influence on recovery and how it might be modulated.

This thesis aims to address these gaps by investigating a method for measuring potential neuroinflammation and exploring a behavioural approach to modifying it. Forty male athletes (aged 16-35 years) were recruited within 14 days of sports-related mTBI, along with 52 age and sex-matched controls. T2-weighted MRI data were collected using a Carr-Purcell-Meiboom-Gill (CPMG) spin-echo sequence. Preprocessing included brain extraction, normalisation, smoothing, and grey matter masking. T2 relaxation times were estimated at each voxel using a monoexponential model in qMRLab. Participants also wore Axivity AX3 wrist accelerometers continuously for 15 days to capture physical activity data. Physical activity data were analysed using the GGIR, compositions and deltacomp packages in R.

The thesis comprises three studies. Studies 1 and 2 assess the utility of MRI T2 relaxometry as a marker of neuroinflammation. Study 1 used a case series approach, statistically comparing individual mTBI participants to the control group via z-tests and voxel-wise z-maps. Results showed elevated T2 relaxation times in acute mTBI, with unique regional patterns per individual and follow-up scans in a subset of participants demonstrated reductions in T2 relaxation times, suggesting recovery. Study 2 complemented this with a group-level analysis. Independent samples t-tests with threshold-free cluster enhancement (TFCE) revealed significantly higher T2 relaxation times in the mTBI group, particularly in frontoparietal regions, as well as a negative correlation between T2 relaxation times and recovery time.

Study 3 shifted the focus to intervention, using compositional data analysis to examine physical activity patterns post-injury. No significant group differences were found, supporting the tolerability of physical activity during early recovery. Although longer sleep duration initially correlated with prolonged recovery, this effect was attenuated after adjusting for baseline symptom severity.

Together, the findings suggest that quantitative T2 relaxometry is a sensitive, non-invasive marker of brain changes potentially reflecting neuroinflammation. The correlation between T2 relaxation times and recovery time may emphasise a dual role of inflammation, beneficial in the acute phase but potentially harmful if prolonged. Study 3 supports the tolerability of physical activity during early recovery and highlights the utility of compositional analysis in understanding movement behaviours. Given the known anti-inflammatory effects of physical activity, its role in recovery may partially stem from modulating neuroinflammation.

These findings support a more individualised approach to mTBI care, integrating objective measures with clinical assessments to identify patients at risk of prolonged inflammation and neurodegeneration. Future research should focus on longitudinal designs with more diverse samples and larger activity datasets to better understand recovery trajectories and optimise intervention timing. As the first to apply quantitative T2 relaxometry and compositional data analysis in mTBI, this thesis offers novel insights into brain changes and physical activity patterns following injury.