Detecting Bipolar Episodes … with Your Phone

Alex Leow, MD, PhD, University of Illinois at Chicago Colleges of Medicine and Engineering

Imagine that your smartphone could alert you to signs of a manic or depressive episode. Soon, it may do just that.

BiAffect is a smartphone application in development for tracking typing kinematics metadata that may indicate mania or depression in people with bipolar disorder. It recently received the top award in the Mood Challenge for ResearchKit, a New Venture Fund program funded by the Robert Wood Johnson Foundation. BiAffect unobtrusively tracks typing metadata on mobile devices, such as typing speed, use of spell-check, how hard keys are pressed, and backspace usage, measures that were found in a pilot study to collectively relate to mood symptoms during manic and depressive episodes.

The app is premised on the clinical observation that those who speak more quickly when experiencing mania would often also type more quickly and less carefully, and the converse that one experiencing a depressive episode would likely type more slowly. To this end, BiAffect researchers noted that backspace keypresses may occur when the smartphone user pauses and determine whether to edit a typo, or whether to accept or overwrite suggestions made by auto-correct rather than to simply keep typing. People in the midst of a manic episode commonly have reduced impulse control, so it is not surprising that their pilot data suggested that some patients tend to blow through the spell-check alerts. On the contrary, during depressive episodes, typing a long message may become laborious, and messages tend to be shorter.

In order to have the technology as unobtrusive as possible and to measure typing in a variety of programs (not only in text messages or in a single-function app), BiAffect integrates through the use of a custom keyboard that replaces the default keyboard.

Not your average mental health program
BiAffect is different than current mental health technologies in significant ways.

At this point, BiAffect is not a brain-training kind of app to improve thinking and memory skills through exercises. Nor is it a medication-management tool or a meditation/mindfulness stress-management program. It doesn’t support development of skills or supply information to your healthcare providers. And it doesn’t do therapy.

Indeed, most of the existing mHealth tools for mental health require substantial user engagement and often rely on self-rated questionnaires, which are known to exhibit social desirability response bias and recall bias, as well as subjectivity. Thus, this field is awaiting disruptive technology and a paradigm shift.

On the other hand, the pervasive use of mobile wireless devices has significantly shaped interpersonal communications. Indeed, as personal smartphone technology advances, people are increasingly interacting with one another via typed communications. Thus, BiAffect researchers seek to convert mobile devices into “fitness trackers” for the brain, by inferring neuropsychological states using keystroke dynamics and other passive sensor information (such as accelerometer).

With technology similar to BiAffect, users may have insights into their mental state much sooner than they would otherwise, despite not having to interact with their phones any differently. Further, to ensure user privacy, the BiAffect app does not track or record the actual content of texts (i.e., it does not track what you type but rather how you type it). Indeed, when people have real-time data indicating deviations from their normal typing kinematics that may suggest early warning signs, they are better positioned to seek help. Earlier intervention tends to yield better, faster results from treatment.

BiAffect’s pilot study demonstrated the feasibility of inferring mood using passively-acquired typing kinematics alone. Researchers hope that similar mobile technologies would hold incredible promise for clinicians in improving diagnostic capabilities by understanding human behavior in people’s “natural environment,” thus leading to early detection, timely intervention, and a better quality of life.

The BiAffect team is currently aiming to launch a large nationwide study later this year in the App Store. When launched, BiAffect will be the first open-science, crowd-sourced scientific study using iPhone to understand mood and cognition in bipolar disorder. If successful, BiAffect will uncover and validate novel real-time virtual biomarkers of “mental-health and brain-fitness footprints.” Moreover, similar technologies can be adapted and applied beyond mood disorders to other neuropsychological disorders like Alzheimer’s and Parkinson’s disease.

Your Turn                                                                                                          

  • In what ways might BiAffect be useful for you or someone you know living with bipolar disorder?
  • What do you imagine the next intersection of mental health and technology will be? What are you hoping technology will do for your mental health?

Further Reading

  1. NIMH, Technology and the Future of Mental Health Treatment
  2. Bakker et al., “Mental Health Smartphone Apps: Review and Evidence-Based Recommendations for Future Developments
  3. Anxiety and Depression Association of America, “Mental Health Apps

Dr. Alex Leow is an Associate Professor in Psychiatry, Bioengineering, and Computer Science at the University of Illinois at Chicago (UIC) and an attending physician at the University of Illinois Hospital & Health Sciences System. With both a PhD in applied mathematics and a medical board certification in psychiatry, Dr. Leow co-founded and co-directs with Dr. Olu Ajilore the CoNECt research team at UIC (, where they study the human brain using interdisciplinary approaches of multi-model brain imaging, non-invasive brain stimulation, Big Data analytics, virtual-reality immersive visualization, and mobile technology.

Facebook Comments