A structural network fingerprint of mild traumatic brain injury: a multi-study synthesis of T1-weighted MRI abnormalities.

BACKGROUND

Mild traumatic brain injury (mTBI) often results in persistent cognitive and somatic deficits despite unremarkable routine neuroimaging. Evidence suggests mTBI affects large-scale neural systems rather than isolated regions, yet structural findings remain heterogeneous across studies.

OBJECTIVE

This study synthesized T1-weighted MRI data into a unified structural network fingerprint (SNF) of mTBI.

METHODS

We analyzed ten peer-reviewed studies identifying regional abnormalities in adult mTBI via voxel-based, volumetry, grey/white-matter probability mapping or tensor-based morphometry. Thirty-five significant regions of interest (ROIs) were extracted and mapped to a standardized anatomical atlas. ROIs were categorized into canonical networks, and we applied co-alteration graph modeling, principal component analysis (PCA), and hierarchical clustering to evaluate network-level convergence.

RESULTS

The SNF identified a core triad of vulnerability: the default mode network (DMN), the limbic/memory system, and thalamic-callosal relay structures. Graph modeling revealed robust clustering among DMN-limbic-thalamic regions. Furthermore, PCA and hierarchical clustering demonstrated that structural alterations strictly align with intrinsic network boundaries, rather than appearing as stochastic damage.

CONCLUSION

mTBI exhibits a reproducible structural signature characterized by DMN and thalamo-limbic involvement. This SNF framework establishes a basis for clinically interpretable biomarkers and computable decision-support tools in concussion care.