ObjectiveTo evaluate the use of neuroimaging safety, diagnoses, and potential treatment of patients with concussion.MethodsThis retrospective study took advantage of a concussion database to analyze neuroimaging in concussion patients of all ages. Details of neuroimaging orders were tracked and categorized as hospital emergency rooms, primary care physicians, neurologists at DENT, or by other specialists. Neuroimaging consisted of MRI and/or CT scans, which were classified as normal or abnormal. Abnormal MRI scans consisted of white matter changes, brain hemorrhage, chiari malformation, cyst arachnoid, hydrocephalus, incidental unrelated finding, or a developmental venous anomaly. Abnormal was further defined as abnormal due to head injury, unrelated to the concussion but unlikely to prolong recovery time, or unrelated to the concussion but may prolong recovery time.ResultsAmong the 835 diagnosed with concussion, 715 (86%) patients ages 1–78 had neuroimaging completed (615 MRI and 422 CT). Among these patients 401 (95%) had a CT order prior to coming to Dent, 319 (80%) from emergency rooms, 64 (16%) from primary care physicians, and 18 (4%) from other physician specialists. The rate of ordering an MRI was 46% greater than the rate of ordering a CT scans, while the rate of discovering an abnormality within MRI scans was 3 times greater than CTs (24.3% vs 7.8%).ConclusionsThere exists a remarkable discrepancy between the rate of ordering neuroimaging in concussion patients (46% more MRI vs CT orders) and the rate at which neuroimaging in these patients discovered brain abnormalities (×2.11 more in MRI vs CT reads). We acknowledge that improvement is required in the length of time in the MRI scanner and cost of MRI technology. However, additional consideration is required in abnormality detection effectiveness, cost efficiency, and radiation safety in balancing the use of MRI and CT technology.

ObjectiveTo compare head kinematics measurements obtained from 6 different head impact sensors utilizing different methods of sensor-to-head fixation. DesignFree-drop impacts (total n = 54) were performed at 3.5 and 5.5 m/s onto to the front, back, side, and top of 2 elderly human cadaveric head-neck specimens: a helmeted (Riddell Revolution Speed) male specimen was dropped onto a NOCSAE testing pad; an un-helmeted female specimen was dropped onto a framed sample of field turf. The specimens were instrumented with an intracranial reference sensor surgically mounted at the approximate head center-of-mass by a rigidly-fixed custom standoff pad, an intra-oral test sensor rigidly fixed to the upper teeth/hard palate by a custom orthodontic appliance, and 4 commercially available head impact sensing systems: X-Patch, Vector mouth guard, HITS (helmeted condition only), and G-Force Tracker (affixed to helmet interior or head band depending on helmet status). Peak linear and rotational head accelerations (PLA and PRA) were compared between each sensor and the intracranial reference sensor using intraclass correlation coefficients (ICC [2, 1]). ResultsAgreement with reference PLA and PRA values differed between sensors, with the greatest agreement observed for the rigidly affixed intraoral sensor (ICC = 0.921, PLA; ICC = 0.810, PRA). Agreement for PLA and PRA, respectively, was: for X-Patch, ICC = 0.638, ICC = 0.155; for Vector mouth guard, ICC = 0.775, ICC = 0.480; for HITS, ICC = 0.662 (PLA only); for G-Force Tracker, ICC = 0.364 (PLA only).

No abstract