Our society faces significant challenges in providing quality health care that is accessible by each person and is sensitive to each person's individual lifestyle and individual health needs. Due to recent advances in sensing technologies that have improved in accuracy, increased in throughput, and reduced in cost, it has become relatively easy to gather high resolution behavioral and individualized health data at scale. The resulting big datasets can be analyzed to understand the link between behavior and health and to design healthy behavior interventions. In this emerging area, however, very few courses are currently available for teaching researchers and practitioners about the foundational principles and best practices behind collecting, storing, analyzing, and using behavior- based sensor data. Teaching these skills can help the next generation of students thrive in the increasingly digital world. The goal of this application is to design online courses that train researchers and practitioners in sensor-based behavioral health. Specifically, we will offer training in responsible conduct, collection and understanding of behavioral sensor data, data exploration and statistical inference, scaling behavioral analysis to massive datasets, and introducing state of the art machine learning and activity learning techniques. The courses will be offered in person to WSU faculty and staff, offered with staff support through MOOCs, and available to the general public from our web page. Course material will be enhanced and driven by specific clinical case studies. Additionally, the courses will be supplemented with actual datasets that students can continue to use beyond the course. This contribution is significant because not only large research groups but even individual investigators can create large data sets that provide valuable, in-the-moment information about human behavior. They need to be able to handle the challenges that arise when working with sensor- based behavior data. Because students will receive hands-on training with actual sensor datasets and analysis tools, they will know how to get the best results from available tools and will be able to interpret the significance of analysis results. Our proposed online course program, called AHA!, builds on the investigators' extensive experience and ongoing collaboration at Washington State University on the development of smart home and mobile health app design, activity recognition, scalable biological data mining, and the use of these technologies for clinical applications. Our approach will be to design online course modules to train individuals in the analysis of behavior-based sensor data using clinical case studies (Aim 1). We will design an educational program that involves students from diverse backgrounds and that is findable, accessible, interoperable, and reusable (Aim 2). Finally, we will conduct a thorough evaluation to monitor success and incrementally improve the program (Aim 3). All of the materials will be designed for continued use beyond the funding period of the program.
This program focuses on the development and dissemination of online courses that train researchers and practitioners in sensor-based behavioral health. Specifically, we will offer training in responsible conduct, collection and understanding of behavioral sensor data, data exploration and statistical inference, scaling behavioral analysis to massive datasets, and introducing state of the art machine learning and activity learning techniques. The courses will be offered in person to Washington State University faculty and staff, offered with staff support through MOOCs, and available to the general public from our web page. Course material will be enhanced and driven by specific clinical case studies. Additionally, the courses will be supplemented with actual datasets that students can continue to use beyond the course.
| Dermody, Gordana; Fritz, Roschelle (2018) A conceptual framework for clinicians working with artificial intelligence and health-assistive Smart Homes. Nurs Inq :e12267 |
| Fritz, Roschelle L; Dermody, Gordana (2018) A nurse-driven method for developing artificial intelligence in ""smart"" homes for aging-in-place. Nurs Outlook : |
| Minor, Bryan; Doppa, Janardhan Rao; Cook, Diane J (2017) Learning Activity Predictors from Sensor Data: Algorithms, Evaluation, and Applications. IEEE Trans Knowl Data Eng 29:2744-2757 |
