Validation of an Algorithmic EEG-Based Diagnostic Tool for Concussion

A study validating  an EEG-based algorithmic concussion assessment tool (Concussion Index; Brainscope, Bethesda, MD) has been published in JAMA Open Network  for evaluating severity of concussions. This diagnostic tool can also be used to an athlete's readiness to return to play and was cleared by the Food and Drug Administration in late 2019. The Concussion Index had a sensitivity of 86%, specificity of 71%, and negative predictive value of 90%. These results support the high accuracy of the Concussion Index in the identification of the likelihood of concussion.

With a physician's order, this algorithmic assessment is done with a disposable headset and handheld device and takes only 20 minutes to administer. Other previously FDA-cleared algorithms are also available on the device including assessing the likelihood of brain bleeds (99% sensitivity) and severity of functional impairment. 

"What this shows us is that for the first time we have a point of care and objective marker that can rapidly identify the likelihood of concussive injury and can be used to follow patients from baseline through recovery," said lead author Dr. Jeffrey Bazarian, Department of Emergency Medicine, University of Rochester School of Medicine. "With its demonstrated accuracy and ease of use in the athletic environment, the Concussion Index has great potential to be incorporated into existing standard assessments of concussion to aid in objective clinical diagnosis and in determination of readiness to return to play."

In the validation study, 580 participating athletes age 13 to 25 years with without concussion were assessed through a variety of methods, including the EEG-based algorithm. Assessments were made within 72 hours of the injury, when returning to play, and 45 days after they returned to play. The methods included electroencephalography (EEG), cognitive testing, and symptom inventories. Of the  580 eligible participants who were analyzed, 207 had concussion and 373 did not. A total of 1,318 evaluations were included in the specificity and sensitivity analyses, including follow-ups. 

"The results of this study are an independent demonstration of the power and reliability of BrainScope's Concussion Index –– as an objective marker in the clinical assessment of concussions at the time of injury and as a reliable indicator of change over time," said Dr. Leslie Prichep, chief scientific officer of BrainScope. "Importantly, the study targeted patients of high school and college age who are at great risk for both short and long-term consequences of concussion, as the brain is still developing."