Susceptibility-Weighted Imaging and Proton Magnetic Resonance Spectroscopy in Assessment of Outcome After Pediatric Traumatic Brain Injury

(A) Conventional GRE (fast imaging with steady-state precession, 500/18, 15° flip angle, 78Hz per pixel, 2 signals acquired, 4mm thick sections) and (B) SWI (3-dimensional fast low-angle shot, 57/40, 20° flip angle, 78Hz per pixel, 64 partitions, 1 signal acquired, 2-mm thick sections reconstructed over 4mm) MR images from the same brain region in a child with TBI illustrating the increased ability of SWI to detect hemorrhagic DAI lesions.

Extent of hemorrhagic lesions compared with GCS, days in coma, or dichotomized outcome. (A) The mean volume and (B) number of hemorrhagic lesions are greater in the more severely injured patients, in those with longer durations of coma, and those with poor outcome at 6 to 12 months. Error bars indicate standard error of the mean. Data based on Tong et al.²¹

Extent of hemorrhagic lesions within individual outcome groups. (A) The mean volume and (B) number of hemorrhagic lesions tend to increase with worsening severity of outcomes. Error bars indicate standard deviations. Abbreviations: MI, mild disability (n=16); MO, moderate disability (n=7); N, normal (n=14); S, severe disability (n=2); V, vegetative state (n=1). Data based on Tong et al.²¹

Examples of hemorrhagic DAI lesions on SWI, visible as areas of hypointense foci. (A) Small shearing injuries at the gray/white matter junction of the left frontal lobe. (B) Typical shearing injuries scattered throughout the hemispheric white matter, including the corpus callosum. (C) Brainstem lesions, usually indicative of more severe injury. (D) Shearing injuries in the cerebellum. The majority of these lesions are not seen or are much smaller on conventional MRI.

Developmental changes in proton MR spectra (simulated-echo acquisition mode; repetition time [TR], 3000ms; echo time [TE], 20ms) from occipital gray matter in 4 different patients at various ages. (A) 4 days, (B) 5 months, (C) 2 years, and (D) 17 years. Note the developmental decrease of myoinositol (3.56ppm) and choline (3.2ppm) and increase of creatine (3.0ppm) and NAA (2.0ppm) with increasing age. Abbreviations: Cho, choline; Cr, creatine; Ins, myoinositol.

(A) T2-weighted MR image and (B) corresponding SWI show hemorrhagic lesions in the body of the corpus callosum and bifrontal extra-axial collections from a 15-year-old adolescent ejected from a car. Patient had a good outcome 12 months after injury. (C) Spectral map shows the 54 voxel (6×9) volume of interest for the MRSI (point-resolved spectroscopy; TR=3000ms; TE=144ms) acquisition. (D) Spectrum from normal-appearing brain in the anterior corpus callosum with decreased NAA (2.0ppm). (E) Spectrum from parietal white matter with normal metabolite ratios.

Shown is the percentage of voxels from normal appearing brain that have NAA ratios below 2 standard deviations from the mean for age and grouped by neurologic outcome assessed at 6 to 12 months postinjury. Approximately, 60% of voxels from normal-appearing brain taken through a transverse section at the level of the corpus callosum in children who have normal or mild long-term outcomes have decreased NAA/creatine ratios early after injury. The percentage of abnormal voxels from normal-appearing brain increases with increasing severity of outcome.