An App To Predict Oncoming Manic Episodes


Researchers estimate that two percent of the global population will be diagnosed with bipolar disorder at some point in their lifetime, with prevalence increasing to over four percent in the US. Bipolar is characterized by abnormal shifts in mood, energy, and activity over time, sometimes manifesting itself as depressing, and other times as a euphoric sensation called mania. Psychiatrists treat the condition with mood stabilizers and for most this is enough, but a team of researchers from the University of Michigan Medical Center are piloting a smartphone app that they hope will help patients predict mood swings before they occur.

The team is being led by Melvin McInnis, MD, a psychiatrist at the University of Michigan Medical Center, who explains “We want to be able to detect (mood swings) well in advance. 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.

Researchers are able to predict oncoming mood swings by analyzing speech characteristics, which are different depending on if a patient is in a state of depression, mania, or in between the two. Depressed patients speak slowly, with longer pauses between words. Manic patients on the other hand speak in quick bursts, and shift topics rapidly. Speech changes are an ideal symptom to use as a benchmark for mood changes because they are some of the first symptoms of a mood change to appear.

The app being piloted is a passive monitoring system that analyzes speech patters while patients are talking on the phone. The app creates a baseline speech pattern for the user, and then monitors changes over time to predict both oncoming depression and oncoming mania.  The app shares the predictions with the patient, as well as their mental health provider.

The team is piloting a beta version of the app with 15 bipolar patients to validate the design and algorithms. So far, the team reports that the app is predicting oncoming mania with a good degree of success, but that the analytics need to be refined to improve the app’s ability to predict oncoming depression.

In May, a prototype of the app was presented at this year’s International Conference on Acoustics, Speech, and Signal Processing in Florence.

Enjoy HIStalk Connect? Sign up for update alerts, or follow us at @HIStalkConnect.

↑ Back to top

Founding Sponsors

Platinum Sponsors