AI Voice Tool Diagnoses Concussions in Seconds

FIU researchers develop AI-powered voice analysis to quickly detect brain injuries

Feb. 3, 2026 at 11:55pm

Got story updates? Submit your updates here. ›

Researchers at Florida International University have developed an AI-powered voice analysis tool that can diagnose concussions in seconds. The tool analyzes subtle changes in a person's voice to detect signs of brain injury, which can be difficult to identify through traditional sideline evaluations. The technology has shown over 90% accuracy in correlating voice changes with confirmed cases of concussion, and could revolutionize how head injuries are assessed in sports and other high-risk professions.

Why it matters

Concussions often go undiagnosed, with an estimated 50% of cases in the U.S. going unreported. This can lead to serious long-term health consequences if players return to action too soon after a head injury. The FIU technology provides a fast, accurate way to identify concussions on the sidelines, helping ensure athletes receive proper medical care and don't risk further damage.

The details

The voice analysis tool developed by FIU professor Christian Poellabauer and his team captures subtle acoustic, phonetic and linguistic changes in a person's speech that can indicate a concussion. By collecting voice samples from hundreds of athletes before and during their seasons, the researchers were able to identify unique "speech biosignatures" associated with confirmed cases of brain trauma. Using AI, the tool can now diagnose concussions with over 90% accuracy, a significant improvement over current sideline assessments that rely on self-reported symptoms. Doctoral candidate Rahmina Rubaiat is working to further simplify the test, allowing trainers to quickly record a single word or sound from each player for baseline and post-incident comparison.

  • The FIU research began around 10 years ago.
  • The voice analysis tool is currently able to diagnose concussions with over 90% accuracy.

The players

Christian Poellabauer

A professor in the Knight Foundation School of Computing and Information Sciences at FIU, who has been studying the correlation between speech biosignatures and traumatic brain injury for around 10 years.

Rahmina Rubaiat

A doctoral candidate at FIU who is working to identify a single word or sound that could be used for both baseline and diagnostic voice samples to simplify the concussion testing process.

Got photos? Submit your photos here. ›

What they’re saying

“The problem is that these tests are not the most accurate.”

— Christian Poellabauer, Professor (Mirage News)

“A severe concussion or mild back-to-back concussions, after a few years, may lead to experiencing some other neurological impairments.”

— Rahmina Rubaiat, Doctoral Candidate (Mirage News)

What’s next

The researchers are working to further simplify the voice analysis test, allowing athletic trainers to easily record and compare baseline and post-incident voice samples to quickly diagnose concussions. They also plan to expand the technology to detect the onset of other neurological diseases like Parkinson's and Alzheimer's.

The takeaway

This AI-powered voice analysis tool has the potential to revolutionize how concussions are diagnosed, especially in sports and other high-risk professions. By providing a fast, accurate way to identify brain injuries, it could help ensure athletes receive proper medical care and prevent long-term neurological damage from repeated head trauma.