A new smartphone app predicts mood in bipolar according to a recent series of studies. The mobile phone app shows promise for identifying early warning signals for users of mood episodes, based on a user’s keyboard behavior, such as typing speed and message length.
The app does not consider the content of messages typed by users on their phones, but rather collects “metadata” about the user’s physical interface with the device. Hypothesizing that at the onset of a manic episode, typing speed will gradually increase (similar to long-standing observations of speed of talking in bipolar patients), the researchers gathered data that suggests that people in the midst of a manic episode do indeed type faster and less accurately than they may otherwise do. Conversely, during a depressive episode, users may type more slowly and send much shorter messages.
If this app predicts mood episodes it might remind a user of their crisis plan or Wellness Recovery Action Plan (WRAP) early enough to prevent serious consequences.
The app is available for free to iPhone users, and an Android version is currently in the works. At present, the app is still under study, and users must opt-in to the research before downloading the app. The researchers hope that enough data can be generated for stronger conclusions to be made, and at some point, more widespread usage of the technique may be useful for millions of people living with mood disorders.
Moodsurfing has been following the news about technology and mental health for some time, and we are always looking at new applications and techniques. A recent post looks at data mining and the promise and potential pitfalls of continuous data collection through personal devices. We advocate for greater user control over the data, believing that this will allow better self-care for those struggling with mood disorders. If you have experience with new apps or other technologies, please share about it in the comments section below.
-Nancy
For More Information
The BiAffect app is available in iTunes app store.