Multicompartmental models and diffusion abnormalities in paediatric mild traumatic brain injury

Area of Science:

• Neuroimaging
• Paediatric Neurology
• Traumatic Brain Injury Research

Background:

• Paediatric mild traumatic brain injury (mTBI) pathophysiology and recovery timelines are poorly understood, relying heavily on subjective reports.
• Adolescence is a known risk factor for prolonged mTBI recovery, but age-specific biomarker impacts remain unclear in large studies.

Purpose of the Study:

• To investigate white matter microstructural changes in paediatric mTBI using diffusion MRI.
• To determine the impact of age-at-injury on diffusion MRI biomarkers in a large, homogeneous sample.
• To evaluate the diagnostic potential of diffusion MRI metrics and advanced modeling in paediatric mTBI.

Main Methods:

• Prospective cohort study involving 203 paediatric patients (8-18 years) with mTBI and 170 age/sex-matched healthy controls.
• Diffusion MRI data acquired acutely (1-11 days) and at 4 months post-injury.
• Analysis using standard diffusion metrics and novel multicompartmental geometric models; machine learning (random forests) applied for diagnostic accuracy.

Main Results:

• Patients showed persistent diffusion abnormalities (increased fractional anisotropy, reduced mean diffusivity) up to 4 months post-injury.
• Multicompartmental models revealed increased intracellular volume and decreased free water fractions, identifying unique white matter pathology.
• Diffusion MRI findings improved diagnostic accuracy by ~10% when combined with clinical data, but were insufficient in isolation.

Conclusions:

• Novel multicompartmental diffusion MRI models are more sensitive to paediatric mTBI pathology than standard methods.
• Diffusion MRI data alone has limitations for achieving high objective diagnostic accuracy in paediatric mTBI due to heterogeneity.
• Evidence of ongoing clinical and physiological recovery at 4 months post-injury suggests a prolonged recovery trajectory.