Collaborative research and fair recognition are the main pillars of EAI conferences. As a researcher, you have the opportunity to connect with the established keynotes or submit and present your study. Conferences like EAI PervasiveHealth 2020 gives us a lot of content to share, and in this blog, we chose two presentations to highlight.
Dana Lewis on artificial pancreas system
Dana Lewis (founder of Open Source Artificial Pancreas System movement) was a keynote at the recent EAI PervasiveHealth 2020. She is the founder of the Open Source Artificial Pancreas System, a community-driven open-source movement that has produced many innovative technological improvements for managing Type I diabetes. She is also an independent healthcare researcher, serving as PI or co-PI on numerous grant-funded projects in leveraging open source diabetes technology to improve diabetes care, and published the first book on automated insulin delivery (‘artificial pancreas’) technology.
In 2013, Dana Lewis reached peak frustration with her inability to hear her continuous glucose monitor (CGM) alarms while sleeping. Open source code enabled her to design her own solutions and eventually iterate and build a DIY hybrid closed loop “artificial pancreas” system. This evolved into the OpenAPS movement, an open and transparent effort to make safe and effective basic Artificial Pancreas System (APS) technology widely available to more quickly reduce the burden of living with Type 1 diabetes.
Dana shared her experiences in designing her own system(s) to help her existing medical devices interoperate, iterating and improving human and computer interactions, and how engaging with patients can help everyone improve healthcare through better research and design. You can watch her presentation #WeAreNotWaiting (and neither should you), where she discusses the OpenAPS movement in this video:
Gym Usage Behavior & Desired Digital Interventions: An Empirical Study presented by Radhakrishnan Meeralakshmi
Understanding individual’s exercise motives, participation patterns and reasons for dropout are essential for designing strategies to help gym-goers with long-term exercise adherence.
In this research, authors Meera Radhakrishnan, Archan Misra, Rajesh Krishna Balan(Singapore Management University), and Youngki Lee ( Seoul National University), derived insights on various exercise-related behaviors of gym-goers, including evidence of a significant number of individuals exhibiting early dropout and also describing their attitudes towards digital technologies for sustained gym participation. By utilizing gym visitation data logs of 6513 individuals over a period of 16 months in a campus gym, they showed the retention & dropout rates of gym-goers. The data indicates that 32% of the people quit their gym activity after initial 1-2 visits and about 65% of the users have less than 10 visits during the study period.
If you are interested in other observations they made, watch their presentation from the Behaviour change session, which also created discussion in conference Slack.
Measuring Self-Esteem with Passive Sensing
In this presentation, authors Bin Morshed, Mehrab (Georgia Tech); Saha, Koustuv (Georgia Tech); De Choudhury, Munmun (Georgia Tech); Abowd, Gregory D (Georgia Tech); Ploetz, Thomas (Georgia Tech), investigated if it’s possible to automatically and scalably predict self-esteem, using passive sensing modalities, available on commodity devices.
Understanding self-esteem can foster adopting preemtive steps to facilitate the psychological and cognitive needs of individuals. Individuals with damaged and lower self-esteem are at a greater risk of psychosocial distress, and maybe vulnerable to the demanding circumstances of day-to-day life.
Therefore, measuring self-esteem is extremely important in many cases. There are multiple ways to do it. The traditional form for measuring self-esteem relies on a survey, which asks experiences of individuals about various things over a long preiod of time. And such practise generates recall bias for the participants since it is hard for individuals to reflect on experiences that happened earlier compared to the time, when they were responding to surveys now how is our approach different.
Smart devices, that we are wearing (phones, smartwatches) are equipped with hardware sensors, such as accelerometer, microphone, gyroscope. Among other things, these sensors can be used for collecting large-scale continuous rich and dense longitudinal data that can be used to infer an individual’s physical and social activities.
Comment from Stephen Schueller – UCI: I’m really interested in this gym usage behavior and how generalizable this behavior / patterns are to lots of health / behavior change areas. The attrition graphs are really similar to what I see in mobile health apps and I’m wondering what lessons I might be able to glean here. I’m also looking forward to making some comparisons in future presentations on my work, as people often ask about the value of apps with high attritions, but I don’t hear people making the same arguments about gyms.
Follow up from Stephen Schueller – UCI: I’d be curious as to your thoughts about lessons that might be able to be taken to other health areas (or health apps). This was great work!
Answer from Meera Radhakrishnan: Hi Stephen, great to know that you are interested in the gym usage behavior as well. Interesting to know that you observed similar behavior among other health apps as well. Based on our studies, one of the main observations was that personalizing the health apps by taking into consideration the individual’s behaviors is very important.
I also feel that in general for all health apps, there should be mechanisms to detect behavior change of individuals, integrate motivational elements and also have personalized interventions at the right time.
Question from Rosa Arriaga: Hi Meera, I also wonder how people that go to gyms compare/differ from people that use exercise apps– did the people you interviewed discuss this?
Answer from Meera Radhakrishnan: Hi Rosa, yes based on our surveys, people reported that the current apps are a one-fit-size solution and they do not get enough benefit. They would like to see apps that are more focused on specific activities (e.g., working out from home vs gyms) and also quantifying the activities at a finer granularity.
Most of the current exercise apps are also only targeted at tracking cardio activities. I think solutions that target other gym exercises (e.g., free weights) are also required.
Would you join a discussion on the presentations? Write a comment to this blog and we will ask the authors to respond.
If you are interested in research articles related to topics discussed on the EAI PervasiveHealth 2020, subscribe to our Pervasive Health and Technology open access journal.
