Play Live Radio
Next Up:
0:00
0:00
Available On Air Stations
Watch Live

Phone App Might Predict Manic Episodes In Bipolar Disorder

There are smartphone apps for monitoring your diet, your drugs, even your heart. And now a Michigan psychiatrist is developing an app he hopes doctors will someday use to predict when a manic episode is imminent in patients with bipolar disorder.

People with the disorder alternate between crushing depression and wild manic episodes that come with the dangerous mix of uncontrollable energy and impaired judgment.

There are drugs that can prevent these episodes and allow people with bipolar disorder to live normal lives, according to Dr. Melvin McInnis, a psychiatrist at the University of Michigan Medical Center. But relapses are common.

Advertisement

"We want to be able to detect that well in advance," McInnis says. "The importance of detecting that well in advance is that they reach a point where their insight is compromised, so they don't feel themselves that anything is wrong."

Early detection would give doctors a chance to adjust a patient's medications and stave off full-blown manic episodes.

McInnis says researchers have known for some time that when people are experiencing a manic or depressive episode, their speech patterns change. Depressed patients tend to speak slowly, with long pauses, whereas people with a full-blown manic attack tend to speak extremely rapidly, jumping from topic to topic.

"It occurred to me a number of years ago that monitoring speech patterns would be a really powerful way to devise some kind of an approach to have the ability to predict when an episode is imminent," says McInnis.

So he and some computer science colleagues invented a smartphone app. The idea is that doctors would give patients the app. The app would record whenever they spoke on the phone. Once a day, the phone would send the recorded speech to a computer in the doctor's office that would analyze it for such qualities as speed, energy and inflection.

Advertisement

Right now the app is being tested with 12 or 15 volunteers who are participating in a longitudinal study of bipolar disorder.

McInnis and his colleagues presented preliminary results at this year's International Conference on Acoustics, Speech and Signal Processing, and so far, things are looking encouraging. McInnis says the software is reasonably good at detecting signs of an impending manic attack. It's not quite as good catching an oncoming depression.

For now, this app is only intended for patients with bipolar disorder, but McInnis thinks that routinely listening for changes in speech could be an important tool for early detection of a variety of diseases.

Copyright 2014 NPR. To see more, visit http://www.npr.org/