Concussion diagnosis is a notoriously tricky science.
Even if an assessment test on the sideline of a football or soccer field deems you to be concussion-free after a blow to the head, that doesn’t necessarily mean you didn’t suffer one, says Michael Singer, CEO of BrainScope, a Bethesda, Maryland-based company that has developed a new technology for detecting signs of a concussion in a person’s brain waves. It has raised approximately $70 million from investors including the U.S. Department of Defense and the NFL.
There is no objective way to detect or assess a concussion, technically known as a mild traumatic brain injury, and BrainScope is trying to change that. Doctors diagnose them based mainly on signs and symptoms. Most people that come into the emergency room after hitting their head also get a computerized tomography (CT) scan to rule out very serious injury. But a CT scan doesn’t detect concussion.
Outside the emergency room, athletic trainers or other medical professionals often use concussion assessment tests, but recent research suggests that those evaluation methods can miss a substantial number of brain injuries. Many people who suffer concussions are never actually evaluated for one.
The consequences of this are complicated and not fully understood. Scientists are still trying to grasp the long-term effects of mild brain trauma, and why some individuals develop related chronic or even neurodegenerative problems because of it (see “What Football Does to the Brain”). We do know that someone who has recently suffered a concussion is at a heightened risk of suffering another brain injury. BrainScope’s technology could eventually be used on the sideline, in the battlefield, or in the emergency room to help clinicians rapidly test for concussions. The test is composed of a disposable headset with sensors that record electrical signals. Placing the headset, recording the signals, and analyzing the data on an Android smartphone application takes just 10 minutes.
The new rapid test is based on a technique called quantitative electroencephalography, or QEEG. Whereas the science of conventional EEG is based on visual inspection of the signal obtained by recording the brain’s electrical activity, QEEG relies on computerized analysis of the signal to detect features that can’t be seen otherwise. BrainScope has developed a set of proprietary algorithms meant to detect sets of abnormalities in the EEG signal associated with the pathology of brain injury.
In the past 10 years, thanks in part to inexpensive computing power and techniques like machine learning, advances in this type of signal processing have been exponential, says Leslie Prichep, BrainScope’s chief scientific officer. “Our ability to characterize electrical signals has just exploded.”
But the fact that an objective concussion diagnostic test doesn’t exist yet makes it uniquely challenging to prove that BrainScope’s test works, says Singer. There’s simply nothing to compare it with. For regulatory purposes, BrainScope and the U.S. Food and Drug Administration have settled on a CT scan as the “gold standard.” The company has published several peer-reviewed studies showing that its technology reliably predicts whether a patient with a head injury will have a positive CT scan.
The first two iterations of the test, though not for sale, have received clearance from the FDA to aid clinicians in deciding whether to administer a CT scan. This is useful; CT scans expose patients to potentially harmful radiation, and are especially risky in children.
But Singer says the company’s goal is to develop a real concussion diagnostic test, capable of giving doctors useful information about the “spectrum” of pathologies associated with concussions that is invisible to a CT scan. The scientific understanding of those pathologies is still limited, but recent advances have been made using new imaging techniques like advanced magnetic resonance imaging (see “Brain Scars Detected in Concussion”).
Crucial to BrainScope’s approach is decades-old work by Prichep when she was co-director of the Brain Research Laboratories at the New York University School of Medicine. In 1988, Prichep and her late husband, E. Roy John, were the first to demonstrate that in normally functioning people, the characteristics of their EEG signal can be described by a set of mathematical equations as a function of age. This means that, using statistics, “we can look at your brain activity and determine how normal it is relative to what we expect for your age,” says Prichep.
The strategy of using such “normative” data to identify abnormalities in a person’s brain waves is not without its skeptics. A number of once-promising biomedical applications based on a similar approach to QEEG have not panned out, says Marc Nuwer, a clinical neurophysiologist and professor of neurology at UCLA School of Medicine. Further, says Nuwer, without a baseline measurement it can’t be known for sure whether someone’s EEG signal is in fact abnormal. The difference could be other things besides concussion, like a medication, a previous head injury, or something else entirely, he says.
Prichep agrees that having a baseline might improve sensitivity. “You may be at your baseline slightly different from normal,” she says. But clinicians are generally not going to have baseline measurements for patients who arrive at the emergency department.
BrainScope is working on a new iteration of its EEG test that is meant to reach beyond what is visible on a CT scan, using signal-processing algorithms designed to take advantage of a newer scientific understanding of concussion pathology. The ability to “check the box” on whether someone is CT positive is a requirement for a useful product, “but it is not the end of the story,” says Singer. “It’s actually the beginning of the story.”
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
Data analytics reveal real business value
Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.
Driving companywide efficiencies with AI
Advanced AI and ML capabilities revolutionize how administrative and operations tasks are done.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.