Radiological-prognostic correlation of diffusion tensor imaging in a mild traumatic brain injury model

Diffusion tensor imaging (DTI) parameters can detect changes in the brain microstructure in mild traumatic brain injury (mTBI). Whether these parameter changes can predict neural functional recovery after mTBI is still relatively unknown. The present study aimed to investigate the radiological-prognostic correlation between these radiological parameters and learning and memory deficits using an in-house constructed rat model of mTBI. We established a rat model of diffuse axonal injury (DAI) at different injury levels, followed by magnetic resonance imaging at 6, 24, and 72 h, and 1, 2 weeks post injury, and randomly selected the rats for analysis of histopathology and learning and memory deficits. DTI parameters and β-amyloid precursor protein (β-APP) levels were obtained to estimate the extent of brain injury and the correlation with the times of crossing the safety platform as measured using a water maze test. The results revealed that fractional anisotropy (FA) was sensitive to axonal integrity. FA values of the corpus callosum in the injury groups decreased at all time points post injury, except in the mild injury group, which recovered to normal levels at 1 and 2 weeks post-injury. The neural function of the mild injury group recovered to normal compared with the normal control group. FA value, β-APP of corpus callosum in different groups at 24 h post injury showed obvious correlation with learning and memory deficits at the recovery stage (r=0.881, r=-0.931). In conclusion, DTI can reflect varying injury states of DAI over time with direct comparison to histopathology and could be used to predict the neural functional recovery at the early stage post-injury in a rat model.