A Longitudinal Study of Closed Head Injury: Neuropsychological Outcome and Structural Analysis using Region of Interest Measurements and Voxel-Based Morphometry

Background: The hippocampus and corpus callosum have been shown to be vulnerable in head injury. Various neuroimaging modalities and quantitative measurement techniques have been employed to investigate pathological changes in these structures. Cognitive and behavioural deficiencies have also been well documented in head injury. Aims: The aim of this research project was to investigate structural changes in the hippocampus and corpus callosum. Two different quantitative methods were used to measure physical changes and neuropsychological assessment was performed to determine cognitive and behavioural deficit. It was also intended to investigate the relationship between structural change and neuropsychology at 1 and 6 months post injury. Method: Forty-seven patients with head injury (ranging from mild to severe) had undergone a battery of neuropsychological tests and an MRI scan at 1 and 6 months post injury. T1-weighted MRI scans were obtained and analysis of hippocampus and corpus callosum was performed using region-of-interest techniques and voxel-based morphometry which also included comparison to 18 healthy volunteers. The patients completed neuropsychological assessment at 1 and 6 months post injury and data obtained was analysed with respect to each assessment and with structural data to determine cognitive decline and correlation with neuroanatomy. Results: Voxel-based morphometry illustrated reduced whole scan signal differences between patients and controls and longitudinal differences between patients which reflected orbitofrontal damage. Using voxel-based morphometry and segmented images, reduced grey matter concentration was found in the patients’ hippocampal region when compared to controls. Data from neuropsychological test scores were related to injury severity as measured by the GCS and PTA. Correlations of cognitive test data with the hippocampus, but not the corpus callosum, were present with voxel-based morphometry providing a greater number of associations than region-of-interest analysis. No longitudinal changes were found in the hippocampus or corpus callosum using region-of-interest methodology or voxel-based morphometry. Conclusions: Decreased grey matter concentration identified with voxel-based morphometry illustrated that structural deficit was present in the head injured patients and does not change between 1 and 6 months. Voxel-based morphometry appears more sensitive for detecting structural changes after head injury than regionof-interest methods. Although the majority of patients had suffered mild head injury, cognitive and neurobehavioural deficits were evident and there were numerous relationships between reduced grey matter concentration and cognitive test scores. The result thus show that subtle effects of diffuse brain damage in the patient group can be detected by neuroimaging and related to cognitive performance.