Susceptibility-weighted MRI in mild traumatic brain injury.

Abstract

To compare the frequency of microbleeds identified by susceptibility-weighted MRI (SWMRI) in patients with mild traumatic brain injury (mTBI) and normal controls, and correlate these findings with neuropsychological tests. Research ethics committee approval and patient written informed consents were obtained. One hundred eleven patients with mTBI without parenchymal hemorrhage on CT and conventional MRI received SWMRI as well as a digit span and continuous performance test. One hundred eleven healthy volunteers without history of traumatic brain injury were enrolled as the control group and received conventional MRI with additional SWMRI study. We analyzed the number and location of microbleeds in both groups. Twenty-six patients with mTBI and 12 control subjects presented microbleeds on SWMRI (p = 0.0197). Sixty microbleeds were found in 26 patients with mTBI and 15 microbleeds in 12 control subjects. The mTBI group showed notably more microbleeds in the cortex/subcortical region (52 microbleeds, 86.7%, vs 3 microbleeds, 20%; p < 0.0001). Conversely, the control group showed more microbleeds in the central brain (9 microbleeds, 60%, vs 3 microbleeds, 5%; p < 0.0001). There was no statistical difference in number of microbleeds in the cerebellum and brainstem (p = 0.2598 and p = 0.4932, respectively). Patients with mTBI who had detected microbleeds had lower digit span scores than the patients with negative SWMRI findings (p = 0.017). Presence of mTBI-related microbleeds showed a neuropsychological defect on short-term memory function, indicating that the presence of microbleeds could be a possible severity biomarker for mTBI. Addition of the SWMRI technique to the MRI protocol for patients with mTBI is recommended.