Cornell University
Search:
more options

Technology as Companion-Coach

This article, written by Amanda Garris, appears in the latest edition of PeriodiCALS, the magazine of Cornell University's College of Agriculture and Life Sciences.

Cell phones have put communication, entertainment, navigation, and restaurant reviews in the palm of our hands. But researchers in the field of information science are making good on the promise to put a psychiatrist and health coach there, too.

“There’s a really good fit between new technologies and designing interventions for mental illness, in general, because these technologies are with us all the time,” says Geri Gay, the Kenneth J. Bissett Professor of Communication and a founding member of the Department of Information Science. “For bipolar disorder in particular, where it’s very important to stay on a schedule of getting up, eating breakfast and making social contact, mobile technologies can actually help track when someone begins to skew into a manic phase—shopping in the middle of the night, checking their phones all night, basically not sleeping—leading to self-destructive behavior.”

The new MoodRhythm app being commercialized by HealthRhythms will provide patients and their doctors a novel way to monitor their well-being. Funded by the European Union and the College of Human Ecology’s Bronfenbrenner Center for Translational Research, the project to develop the app brought together a team of researchers and clinicians led by information science faculty Gay and Tanzeem Choudhury, postdoctoral associates Mark Matthews and Elizabeth Murnane, Ph.D. ’17, and Ellen Frank, distinguished professor emeritus of psychiatry and distinguished professor of psychology at the University of Pittsburgh.

The app combines patients’ self-assessments with passive monitoring of behavior to provide feedback on how well they are meeting their targets for physical activity, social interaction and sleep—the guardrails that mark the road to health. Physical activity is tracked using the phone’s accelerometer, the number and duration of social interactions are estimated through privacy-sensitive monitoring of conversations detected by the built-in microphone, and sleep duration is logged by the light sensor. Users get a weekly report that provides feedback on how well they maintained their target rhythms, but in a format that looks more like a game than a medical tracker, with animated visualizations of rhythm and awards and badges for adhering to the schedule.

“It’s essential to design for delight. If an app is pleasurable, people tend to spend more time with it,” Gay says. “We’ve found in general that simple elements that make things fun or quick games to play give a little spike in people’s happiness levels.”

Patients provided crucial design insights to strike a balance between vigilance and flexibility. Some changes to the app based on patient input included a “cheat day” and a reset of tracking every seven days to buffer against discouragement. Perhaps the most crucial feedback was the high degree of satisfaction that patients expressed with the app and, in particular, the passive sensing of behaviors.

“People have told us the idea of intelligent tracking is extremely appealing. The patients really see a need for it,” Murnane says. “It could help them preempt a manic episode before they’d normally even notice they are starting down the behavioral path that leads to symptoms.”

Other mobile health technologies developed in the Interaction Design Lab directed by Gay have addressed a range of conditions, from a mobile phone game to encourage mindful eating in adolescents to Nutriphone, a smartphone-based micronutrient testing system currently under commercialization and approval by the FDA. But Gay sees a great potential for mobile technologies to improve the treatment of other disorders where self-assessments are tough, including depression and chronic pain. The latter is the subject of a project funded by the National Science Foundation to evaluate if behaviors logged by cell phone sensors can serve as biomarkers of chronic pain.

“Chronic pain is quite an issue in the country, and we’ve been looking at alternatives to pain-controlling drugs, which have created another rampant problem,” says Gay, who collaborates on the grant with Choudhury and Deborah Estrin, professor of computer science at Cornell Tech and Weill Cornell Medical Center. “It’s complex to even measure pain because it is so different for each person.”

That’s why the first challenge of the project was to discover how people prefer to answer the question, “How much does it hurt?” Is a person’s pain a seven on a scale of one to ten, a stapler to the spine, or a grimacing facial expression? They tested nine different ways to report pain and learned that no one-size-fits-all solution for pain assessment exists. Some people preferred a numerical scale, while others responded better to more abstract representations of faces and landscapes. It’s one reason why the team focuses not only on refining self-reporting but also on automated data collection to give doctors and patients accurate information at a granular level, for example, using the accelerometer to record how fast a patient rises from a chair each morning.

“Our intention is that these technologies become empowering, allowing people to take control over their own health and well-being,” Gay says. “Through these devices, we are being tracked all the time. Giving individuals control over their own data is important to me.”

Although the most recent applications developed in the Interaction Design Lab are for chronic, difficult medical conditions, cell phone data is giving information scientists unprecedented insight into our natural biological clocks and what happens when we disrupt them.

“Everybody probably has a sense of their chronotype: You know if you prefer to wake up early or stay out late. You might notice a regular crash after lunch when you can’t get any work done,” Murnane says. “Pretty much every single biological process—our digestion, concentration, mood, athletic performance and even sensitivity to pain—exhibits some sort of 24-hour, circadian rhythm with a genetic basis.”

The problem arises when your circadian clock is not in sync with your alarm clock, and it’s an epidemic. If you wake up earlier in the morning on weekdays and sleep in on weekends—an estimated 80 percent of people do—you are likely suffering from a form of jet lag. Unlike the temporary jet lag from travel across time zones, the resulting “social jet lag” can have serious consequences for health and performance. Alertness, Murnane notes, is the “lynchpin of cognitive performance,” and your cell phone can detect its ups and downs.

“The tech-mediated markers—behaviors you can capture passively through the phone—can be quite revealing,” Murnane says. “We found strong correlations with social jet lag as well as other indicators of circadian disruption like sleep inertia. Similarly, when we looked at relationships between technology use and daily performance patterns, we saw individuals using their phones in very different ways, in accordance with their natural alertness rhythms. The types of apps you use, when, and how provide an informative signal about your latent biology.”

Murnane’s work has shown that there is natural variation in biological clocks, and late rising—traditionally stigmatized as evidence of laziness—is completely natural for some. Based on insights from circadian computing, she advocates for a more rational approach for scheduling our lives to take into account variation in biological clocks.

“I’m particularly excited about the opportunity for novel circadian-aware technologies that better align with our own biology to stabilize or even preempt disruptions,” Murnane says. “Such tools could help you embrace your own rhythms to the best you can.”

For example, teenagers’ biological clocks are shifted later than adults—lateness peaks around age 20—which suggests that delaying high school start times would better prime students for learning. If a better match can’t be made, Murnane notes that interventions like light therapy, well-timed naps, or small shifts in sleep schedule can mitigate the effects of disruption. Future apps could help us work in harmony with our circadian clocks.

“I like to imagine technology as our partner,” Murnane says. “It is evolving into a kind of companion-coach, especially in the context of health. Technology just needs to become a little savvier to support us in a way that augments what we could do on our own.”