Although brain MRI is the standard of care in dementia assessments, its application to TBI is not routine even though TBI itself is an independent risk factor for cognitive decline and dementia.

The purpose of this work was to apply automated quantitative volumetric MR neuroimaging to characterizing specific patterns of brain volume loss in a cognitive neurology cohort of individuals with traumatic brain injury (TBI).

As part of an IRB approved study, 21 cognitive neurology patients from UCLA with history of TBI were retrospectively evaluated. Patients were initially evaluated in a Memory Disorders program for a history of cognitive decline.  All subjects received a sagittal T1 weighted volumetric acquisition on a Siemens 3T MR scanner. Each scan was interpreted visually by a board certified neuroradiologist. All volumetric scans were analyzed by an FDA cleared volumetric algorithm, Neuroreader, that computed Z-scores and percentiles compared to a normative database.

Subjects with history of TBI in our sample had a mean age of 65.5 ± 13.9 years, were 43% women, and had the following mechanisms of TBI: 52.4% ground level falls, 23.8% motor vehicle collisions, 14.3% direct blows to the head, and 9.5% blast injury. There was increased brainstem, ventral diencephalon, and pallidum volume loss compared to age, gender, and head size matched controls. Mean z-scores were -2.05 standard deviations (Brainstem), -5.92 (Ventral Diencephalon), -5.74 (Pallidum). By contrast, the hippocampus was -1.01 standard deviations below the mean.

Quantitative volumetric MR neuroimaging shows volume loss specific to TBI in the brainstem, ventral diencephalon, and pallidum. These areas are distinct from hippocampal volume loss reported in Alzheimer’s dementia and suggest a central axis of pathology in TBI. Quantified volumes on MR imaging may be useful in identifying TBI related brain atrophy.