Add another item to the now familiar category of “there’s an app for that” — bipolar disorder self-care. A new smartphone app can detect and monitor subtle shifts in a person's voice during phone conversations and so may identify the early warning signs of potentially self-destructive mood swings in patients suffering from manic-depressive illness.

Though hopeful, the researchers from University of Michigan who developed the app say it still needs more testing before it may be adopted for daily use. "These pilot study results give us preliminary proof of the concept that we can detect mood states in regular phone calls by analyzing broad features and properties of speech, without violating the privacy of those conversations," said Dr. Zahi Karam, a postdoctoral fellow and specialist in machine learning and speech analysis.

The app, which runs in the background on smartphones with Android operating systems, records only one side of a conversation (and therefore complies with laws about recording calls). Automatically, it logs patients' voice patterns during any calls made as well as during weekly conversations with a member of the patient's care team. Next, a computer program analyzes characteristics of the sounds — including all the silences — embedded within each one-sided conversation. Meanwhile, the recordings themselves are encrypted; the research team can see only the results of computer analysis of the recordings, which are stored in secure servers also complying with patient privacy laws.

All of the original six patients to test the product have a rapid-cycling form of bipolar disorder type one, which means they frequently alternate between serious episodes of depression and mania. Analysis of their voice from everyday conversations detected mood swings, the researchers found, when combined with weekly assessments with a trained clinician who had established a benchmark for the patient's mood. The researchers noted they will improve the app’s technology and algorithms over time, boosting its ability to detect mood states, as the software is fed more conversations (and so more data) from volunteer patients. Currently, 60 patients receiving treatment at the nation's first center devoted to depression and related disorders are using the app and adding to its databank.

"This is tremendously exciting," said Dr. Melvin McInnis, a psychiatrist and bipolar specialist. "The ability to predict mood changes with sufficient advance time to intervene would be an enormously valuable biomarker for bipolar disorder." The University has applied for patent protection for the intellectual property involved. Karam, McInnis, and their colleagues presented their findings, funded by the National Institute of Mental Health, at the International Conference on Acoustics, Speech and Signal Processing in Florence, Italy.