Play Live Radio
Next Up:
0:00
0:00
0:00 0:00
Available On Air Stations
Contact 90.5 WESA with a story idea or news tip: news@wesa.fm

Pittsburgh Startup Wants To Turn Wearable Tech Devices Into Relapse Prediction And Prevention Tools

Mark Lennihan
/
AP Photo
Pittsburgh startup Behaivior's software is designed for wearable devices like Fitbit's Alta HR.

Wearable technology like Fitbits and Apple Watches can measure physiological signals and track a user’s location.

One Pittsburgh startup wants to use these features to turn such devices into tools for predicting and preventing relapse in recovering drug users.

The craving states of recovering drug users aren’t just mental: they’re reflected physically in signals like heart-rate, body temperature and respiration. And certain location cues can be tied to an increased risk of relapse, said Ellie Gordon, co-founder of Behaivior.

“So, if someone isn’t attending their weekly Narcotics Anonymous meetings or if they’re near one of their trigger locations, like where they used to use or where they used to purchase [we can take action],” said Gordon.

Behaivior, currently based with the accelerator AlphaLab, is creating AI software for wearables that read into these factors in the user.

Gordon said that when the AI detects risky patterns, it takes actions pre-determined by the user, like contacting someone in their support network and/or sending them a personalized message.

“The beauty of our technology is that interventions can be completely customized in advance to do whatever people know will help the most,” said Gordon. “We will [also] be able to detect overdoses and give quicker access to life-saving medications like Narcan. So our goal is to be a guardian angel on your wrist.”

Behaivior is currently testing with patients in recovery from opioid addiction. Baseline measurements for each patient are compared to those taken at points when patients self-report craving states.

This is necessary because in different people, different factors or combinations of factors will be more predictive of a risky state than others, said Gordon.

Gordon said that although Behaivior is currently focusing on opioid addiction, it’s not the only area they want to venture into.

“[We want] to create a general behavioral pattern recognition model using AI,” said Gordon. “We plan on expanding to help people struggling with other addictions or conditions like PTSD and overeating or any [problematic] behaviors they want to change.”

Behaivior is currently one of 59 teams (and one of seven from Pittsburgh) still competing in IBM’s AI XPRIZE competition, which will award $5 million prizes in 2020.