Magnetic resonance imaging of traumatic brain injury: relationship of T2 SE and T2*GE to clinical severity and outcome

Abstract

Primary objectives: To evaluate (1) the sensitivity of magnetic resonance imaging (MRI) T2* weighted gradient echo (GE) vs T2 weighted spin echo (SE) technology for lesion detection in traumatic brain injury (TBI) and (2) the relationship of lesion patterns to acute clinical severity and 1 year post-injury outcome measures.Research design: Comparative analysis.Methods and procedures: Forty-three acute rehabilitation patients with TBI were imaged utilizing T2 SE and T2* GE techniques an average of 26 days post-injury. Acute clinical severity measures, including Glasgow Coma Scale (GCS), time to follow commands (TFC) and post-traumatic amnesia (PTA) were abstracted from medical records. One-year post-injury outcome measures including Glasgow Outcome Scale (GOS), Disability Rating Scale (DRS) and the Craig Handicap Assessment and Reporting Technique—Short Form (CHART-SF) were collected as part of a comprehensive annual follow-up.Main outcomes and results: In comparison to T2 SE, T2* GE more frequently detected lesions in each of the cortical (p < 0.0001), white matter (p < 0.001), central grey (p < 0.001) and brainstem (p < 0.01) regions and in each of the frontal (p < 0.0001), temporal (p < 0.0001), parietal (p < 0.001) and occipital (p < 0.0001) lobes. With regards to acute clinical severity measures, T2* GE findings were the best predictors of GCS and the only significant predictors of PTA, while T2 SE findings were better predictors of TFC. For 1 year post-injury outcome measures, multivariate regression models utilizing T2 SE and T2* GE findings in combination were the best predictors of DRS and GOS and T2 SE findings alone were the best predictors of CHART-SF.Conclusions: This study demonstrates the enhanced sensitivity of T2* GE for detecting haemorrhagic lesions associated with TBI and supports a complimentary role for both T2 SE and T2* GE weighted imaging in characterizing injury severity and predicting longer-term outcomes.