At the American Diabetes Association (ADA) 80th scientific sessions last week, Dr. Ashutosh Subharwal, Department Chair and Director of Department of Electrical and Computer Engineering at Rice University, put together a compelling presentation that showed the benefits of using smartphone sensors to measure behavior-biology pathways and use those findings to assess, treat and improve healthcare outcomes for depressed patients living with diabetes.

Scaleable Health Labs believes that there should be a bio-behavioral sensing layer to healthcare using simultaneous, non-invasive and accurate measures to provide clinicians with data to better help their patients. They feel using quantitative data from a sensor-based automatic measurement will be more useful than asking patients questions and having their answers be based on their own perception. For example, when asking patients how often they exercise, their answer may vary drastically from the data from their sensor.

The presentation focused on 3 areas: Mobile bio-imaging, behavioral sensing, and data science for health.

Dr. Subharwal proposed two questions:

1. Can we track depressive/anxiety states?
2. Can we measure loneliness?

Can We Track Depressive/Anxiety States?

Depression is a common comorbidity of diabetes and is often undetected and untreated. A study explored this in adults and adolescents, using a tool called SOLVD: Smartphone and Online Usage as based eValuation for Depression, by way of a smartphone as a wearable for tracking depression.

The two clinical pilots for SOLVD consisted of the following:

• Bi-weekly clinical visits.
• Logging feelings in a MoodReminder Module.
• The MobileLogger Module, which has the sensor logging social use (phone calls/texts), mobility (GPS/steps/accelerometer) and phone usage (screen time, screen light, etc.), all while being respectful to keep any conversations private.
• A new parent app for the teenage pilot that used the parent’s feedback as a sensor to measure their children’s mental well-being.

Using this combinational sensor data allowed clinicians to track who, when and for how long the patient was speaking with or texting an individual. They were able to track where the patient was going and the duration of time spent at each location. There were also many other extracted features from the smartphones related to communications, mobility and sleep collected daily, as listed in the chart below.

Key Findings

This method proved to be a useful way to continuously track a patient’s mental state. The patients did not find it intrusive and were willing to be tracked. They saw a strong correlation between the daily self-reported moods and different diagnostic questionnaires in both teens and adults. Also, when patients had fewer phone calls/text messages and shorter frequency of these exchanges, it was predictive of higher depression symptoms. Additionally, as the number of steps walked decreased, there was an increase in the participant’s depressive state.

In both studies, there was a correlation between the data collected from the smartphone and the patient’s psychometric scores, and a noticeably stronger correlation in the moderate to severely depressed participants. The data indicate that the more depressed a patient was, the less mobile and social the person became. This information can help providers to better assess and treat their patients.

Can We Measure Loneliness?

Sociability is crucial to our overall well-being and lack of social encounters are indicators of loneliness. The traditional measures of sociability are often questionnaires, patient self-tracking, the UCLA loneliness scale but all of these require participant effort and many times the report lacks enough detail to draw any conclusions.

SocialSense, an in-person social network (IPSN) is able to track real-life, in-person interactions through audio data. This tracking device is able to detect conversations, detect social scenes and context as well as turn-taking behaviors, with no content analysis to respect participant’s privacy.

The Sociability Clinical Pilot at Baylor College of Medicine (emailed waiting for confirmation) spent 1 week audio-tracking their participants, using the daily smartphone app sensor features discussed above, along with patient baseline psychometric measures.

Key Findings

A decrease in sociability was seen among patients with depression, including fewer longer conversations and fewer social contacts. The SocialSense reports were consistent with the self-reports. SocialSense was also able to detect audio self-talk conversations amongst patients with psychosis.

Conclusions

• Most patients are willing to be monitored via technology (>80% adherence).
• These tracking studies are among the first of their kind to study adolescents and adults who suffer from depression. They are also the first to use the new tool for psychiatry, the parent app.
• Data from the participant’s phone sensor and usage features correlated with symptoms of depression, which was even more pronounced in the moderate to severely depressed patients.
• The data we can get from wearables can help better evaluate a patient’s mental well-being and develop the most appropriate solutions.

What are your thoughts on the subject? How would you feel about your activities being tracked for health purposes?

Post Views: 4

Read more about ADA2020, American Diabetes Association (ADA), depression, diabetes burnout, exercise, mental health, Technology.