A new mouthguard has been shown in lab testing to be 99 percent accurate in predicting serious brain injury after near-concussive force, a new study reports, holding profound implications for sports like football and hockey where hard hits are built into the game.

As the science grows behind concussions and traumatic brain injuries, coaches, parents, athletic programs, and the players themselves are beginning to appreciate the risks involved with repeated blows to the head. As the skull gets bashed over and over, the spongy brain slams into the bone’s inner walls. And while a pool of protective fluid offers a mild cushion, the organ can only take so much pounding. Eventually it begins to degrade.

The new study comes from Stanford University researchers who wanted to learn how effective an in-mouth sensor could be at measuring the force rippling down from a rattling brain. Their experiments on crash-test dummies offered some promise, but the mouthguard idea seemed to come with natural limitations. For one, they found a player could easily throw the mouthguard on the ground and it would register a pretty substantial force, as if it had actually been sitting between gritted teeth, mid-tackle.

The team needed a better model. It had to be sensitive enough to pick up the force of a hit, but precise enough to avoid false positives. Their answer was infrared light. When teeth come in contact with the mouthguard, alerting the team the device is actually being used, the teeth’s natural properties of absorbing and scattering infrared offer a glimpse into how much force registers at the moment of impact.

Tinkering with the model will hopefully let the research team collect large data sets on actual players, in-game. Rather than isolate one hit after another, they could use the information to generate a snapshot of injury — understanding, for instance, how a primary injury floods into each subsequent blow. “We do know that sustaining a second injury right after the first injury will exacerbate the trauma, so detecting that injury is critical,” said Lyndia Wu, a bioengineering doctoral student and the study’s lead author, in a statement.

The problem is, too few players vocalize their injuries, either because they don’t realize they’re hurt or because they don’t want to let their team down. To be safe, athletic programs need hard data. They need G-force “danger zones” that immediately alert coaches when smashing hits send players’ heads spinning. “A player typically shakes it off, thinking he will be fine, without telling the coaches or trainers,” Wu said. “Eventually, we hope to have a device that is able to screen for injury in real time.”

The data surrounding concussions may be misleading. Over the last seven years, diagnoses have doubled among high school sports, but not because there are twice the concussions; we’re just better at noticing them. (For now, let’s ignore the tired question: If a player gets a concussion, and no one diagnoses it, is still it a concussion?) Even at the pro level, the NFL has taken concrete, if tiny, steps toward concussion reduction, the least babyish of which being a small fine for players who lead with their head during a tackle. However, dozens of ex-athletes have sued the league claiming they never knew their day jobs would turn their brains into applesauce. The litigation is still pending.

While research is still underway to find out what causes such traumatic, long-lasting injuries, the Stanford team is doing their best to prevent them in the first place — not to mention those injuries that result from the mundane, everyday hazards like trips and falls. For that, the researchers are relying on a resource fairly close at hand.

“Our football team has been extremely cooperative and interested in helping solve this problem,” said David Camarillo, assistant professor of bioengineering. “Football players willingly put themselves at risk at a well-defined point and time in space for us to carry out our research in this ‘lab.’ What we are learning from them will help lead to technologies that will one day make bike-riding and driving in your car safer, too.”

Source: Wu L, Zarnescu L, Nangia V, Cam B, Camarillo D. A Head Impact Detection System Using SVM Classification and Proximity Sensing in an Instrumented Mouthguard. Biomedical Engineering. 2014.