Developing Algorithm at MIT

MIT’s Lincoln Laboratory (MIT LL) www.ll.mit.edu, a federally funded R&D center, is developing a computer algorithm to identify vocal biomarkers to help diagnose mild traumatic brain injuries or concussions. The U.S Army Medical Materiel Agency (USAMMA), has partnered with the MIT LL to help evaluate, test, and validate the algorithm.

The algorithm has a successful track record in assessing mental or cognitive state from speech. So far, researchers have used the algorithm to identify a variety of cognitive changes, including changes associated with preclinical mild TBI and depression.

For example, people with mental or cognitive changes may elongate their syllables and vowel sounds, change the coordination of their tongue and lip movement, struggle with pronouncing phrases that require complex facial muscle movements to sound out. Also, their voice quality may fluctuate or change.

Pilot studies using the same technology have also demonstrated the potential to detect and objectively measure symptoms in a range of conditions including not only depression, mild traumatic brain injury, concussion, cognitive impairment, but also for Parkinson’s disease.

Growing evidence suggests that a much wider range of mental and physical health conditions impact the major systems involved in speech production and produce distinct vocal biomarkers that may make this technology be useful far beyond initial demonstrations.

MIT has reached an exclusive licensing agreement with Sonde Health www.sondehealth.com. Sonde’s focus areas include mental health conditions as well as a respiratory and cardiovascular conditions where remote passive monitoring could prove to be helpful.

Sonde a leader in the field of health monitoring audio analysis technology along with other experts are developing a portable device similar to a smart phone to host the algorithm to allow it to be used as a medical diagnostic device or decision aid. The goal is for the device to be ready to be cleared by FDA by 2018.