Mobile Technologies Core

Empowering mental health research for improved treatment and prevention 

Mobile technologies have the potential to transform mental health research and precision treatments. Combining mobile apps and devices such as wearables enables the collection of mental health health assessments alongside other objective physiological signals and social parameters, allowing long-term research to be done at scale with greater participation. These technologies also hold the promise to help develop more dynamic, real-time, individualized mental health treatments.

The Mobile Technologies Core facilitates effective, rigorous, equitable, and reproducible mental health research using mobile technologies. Our aim is to empower researchers across the University of Michigan to transform the future of mental health treatment and prevention by overcoming the barriers of incorporating mobile technologies in their studies. The Mobile Technologies Core is partnering across the University to build the infrastructure necessary to make this a reality.

Our Services

Our dedicated faculty and staff provide expert guidance and best practices for Eisenberg Family Depression Center members interested in using mobile technologies in their research

  • Research Proposal Ideation and Development
  • Selection and Use of Mobile Technologies: Devices, Apps, & Platforms
  • Mobile Data Processing and Analysis Support
  • Connection, Dissemination, and Education

Questions and topics to get started:

⟶ Why use mobile technologies in your research

Mobile health technologies offer advantages like the ability to assess various measures together, both objectively (e.g. wearable sensors) and with self-report (e.g. app acquired ecological momentary assessment).   
Data collection methods of mobile technologies are unique compared to traditional research tools, and can acquire long-term data in free-living conditions, often with minimal effort on the part of the study participant. Meta-data is continually collected alongside the research parameters of interest which enrich the data with contextual information.

Mobile technology measurements that may benefit a study include:

⟶ Elements of research with mobile technologies

The Mobile Technologies Core will provide support and share expertise for the following technologies and technology applications. 

  • Off-the-shelf wearables and other mobile monitoring devices
  • 3rd party apps and websites
  • Smartphones
    • Used to deploy apps
    • Contain native sensors to collect data (microphone, gyroscope, etc). 
  • Artificial Intelligence (AI)
    • Automate participant data collection and deploy interventions through apps
      • ie: Chatbot
    • Al algorithms 
      • Used by mobile technologies at multiple levels, including: 
        • AI analysis of wearable signals to derive meaningful quantities (i.e. total sleep times)
        • AI analysis of health metrics (i.e. total sleep time) obtained from wearables combined with data derived from apps (i.e. self-report mood levels) along with other metadata (big data analysis)
  • Telemedicine
    • Studies using wearables and telemedicine delivery of care.
Faculty interested in developing apps for commercialization should visit:  UM Innovation Partners 



Sleep is one, of many, specific measurements that mobile technologies can help collect. We will be covering more topics on this page.


⟶ Why track ambulatory sleep for your project?

Paradox of sleep medicine and research   

We know that sleep deprivation and sleep disorders are problematic not due to singular exposure, but due to nightly disturbances. However, our gold standard to measure sleep is the polysomnogram (PSG) which is a great tool for obstructive sleep apnea (OSA), but given extensive monitoring of multiple physiological parameters, PSG is impractical beyond 1-2 nights.

Objective sleep tracking over time can utilize FDA cleared, traditional actigraphy devices but multiple limitations (cost, owned by the health system/research team, no integration with other digital tools, no real time data acquisition/analysis, not acceptable/desirable by participants) →

LIMITED RECORDING DURATION, PROJECTS NOT SCALABLE

Problem with only using self-report sleep measures
Inaccurate (and direction + magnitude based on demographics and comorbid sleep disorders)
Incomplete
Burdensome
Capture only one or a few aspects of sleep (relevant associations missed)

MULTIPLE COMPONENTS DEFINE SLEEP HEALTH!

Sleep health is not fully encompassed by sleep duration! Simultaneous consideration of the variables has been found to best predict outcomes like mortality. Additionally, daily variation of sleep parameters around the mean is associated with psychiatric, metabolic, cardiovascular consequences.

Therefore, longitudinal, passive recording of objective sleep parameters augmented by self-report data is required to fully understand the role of sleep in health and disease

Wearable sleep tracking technologies

What are they?
Sleep capabilities are frequently included in smartwatches and fitness trackers but are emerging as ring devices and headbands as well. At a minimum, these devices record motion and heart rate with accelerometers integrated into microelectromechanical systems (MEMS) and photoplethysmography. Algorithms are applied to recorded signals then compute activity, cardiovascular, respiratory, and sleep-related metrics. Other sensors, depending on the device, may include electroencephalogram, skin conductance, and temperature.

PSG versus wearable sleep tracking technologies
While incredibly valuable to objectively measure sleep passively and over time, investigators should understand the inherent differences between PSG and wearable sleep tracking devices.


From de Zambotti M, Menghini L, Cellini N, Goldstein C, Baker F. Performance of Consumer Wearable Sleep Technology in Encyclopedia of Sleep and Circadian Rhythms, 2e

Accuracy
Wearable sleep tracking technologies are assessed against PSG to determine their sleep tracking capabilities. In general, motion and heart rate-based wearable sleep tracking technologies are highly accurate in identifying sleep during the time in bed period, but often misclassify non-moving wakefulness as sleep, which can result in overestimation of total sleep time. Ability to identify sleep stages (light sleep, deep sleep, REM sleep; see below) are currently unclear.


Meet Our Team

Core Faculty Lead
Core Staff

Get in Touch

 We’re excited to work with you on your project. Submit the following form to connect with our staff and faculty lead and learn more about how we can help. The Mobile Technologies Core is continuing to evolve and develop new resources. By signing up you will be among the first to know about the latest updates.