PDS Workshop Programme

Monday 5th September

20.00 – 21.30        Welcome Reception, Hotel Edge, Lobby level

Tuesday 6th September

09.00 Welcome

09.05 – 09.20Corpore Sano Centre”, Dag Johansen, Computer Science, UiT

09.20 – 09.50The Tromsø population study: present opportunities and future challenges”, Sameline Grimsgaard, Faculty of Health Sciences, UiT

The Tromsø Study is an epidemiologic and prospective study of lifestyle, health and chronic diseases, and a resource for public health surveillance. The study was initiated in 1974, and six surveys have been carried out, referred to as Tromsø 1-6. Currently, Tromsø 7 is ongoing, and this talk will present experiences from collecting and analyzing data in these studies, but with emphasis on next generation epidemiological studies where pervasive technologies like smartphones and sensor technology might play a pivotal role.

09.50 – 10.15 Quantifying Actions and Player Health in Professional Soccer”, Svein Arne Pettersen, School of Sport Sciences, UiT

Elite athletes are currently monitored, measured, tracked, and tested frequently to ensure that their training load and recovery are balanced. We are in particular interested in improving both the personalized training load to optimize performance and our understanding of the risk factors and mechanisms of sport injury and illness in order to improve training regimens and prevent injuries. In this talk, we will present initial experience from using quantified data in the elite soccer domain, where we are taking advantage of new digital technology to record data.

10.15– 10.40  ”Type 1 diabetes = life-long logging of health data”, Gunnar Hartvigsen, Computer Science , UiT

People with Type 1 diabetes continuously need to monitor their own health. Blood glucose regulation depends on correct medication (insulin), nutrition and physical activity. Through the use of self-management tools people with diabetes are able to keep track on how they are doing. Analysis of previously recorded data enable these patients to better predict their blood glucose development in the near future. If the recorded data are made available on the Internet, the data can be used in group-based applications for motivational purposes. In Tromsø, we are currently working on a disease surveillance system based on reuse of data from people with diabetes. The presentation will show how data from people with diabetes can be used to increase the quality of life for this patient group.

10.40 – 11.15 Break (refreshments)

11.15 – 11.45    “Quantifying dietary intake to support increasingly precise evidence-based guidance on dietary requirements to avoid chronic disease”, Patricia A. Cassano, Div. of Nutritional Sciences, Cornell, Ithaca

The measurement of diet is a key challenge in studying the role of nutrition in chronic disease and morbidity. Epidemiologic studies currently rely on self-reported, paper and pencil dietary intake assessment methods, which are limited by poor accuracy, low precision and high participant burden. Novel methods to assess dietary intake that address these weaknesses by incorporating mobile technologies are in demand, as reflected in current calls for methods development from the US National Institutes of Health (NIH), the recently released US National Nutrition Roadmap, and the newly formed Office of Nutrition Research within the US NIH. The development of improved measurement is expected to lead to better understanding of the role of dietary patterns in maintaining healthy lifespan and preventing chronic disease, and to efforts to develop the area of precision nutrition (dietary requirements/guidance tailored to population subgroups) by studying variability in the individual response to diet and incorporating –omics technology.

11.45 – 12.15Next-generation Video Surveillance: from soccer analysis to disease detection“, Pål Halvorsen, Simula Research Labs, Norway

Video systems have developed from closed local systems to systems used online everywhere. In Corpore Sano, we have automated several labor intensive tasks through use of video. In this talk, we describe how we have assisted the move from traditional pen and paper used earlier by soccer coaches to give feedback to their players during and after practice sessions and matches, to a fully automated video system orthogonal to many of the existing game analysis systems. Furthermore, based on knowledge from the sports domain, we have developed a video disease detection system for the gastrointestinal tract, and we will present the system and some initial results.

12.15 – 12.45      “Visual lifelogging”, Cathal Gurrin, Dublin City University, Ireland

Visual lifelogging has been the subject of research for more than a decade, yet, for a number of reasons, it has not yet lived up to expectations. In this talk, we will explore the potential of visual lifelogging, identify what is needed to bring it to mainstream use, and we will review notable research to date on the area, including a decade of research prototyping at DCU. During the talk, we will explore novel applications in the healthcare domain and introduce initiatives such as the NTCIR-Lifelog comparative benchmarking exercise. Finally, we will wrap up by considering the societal impact of a world in which visual lifelogging is a normative activity.

12.45 – 13.45      LUNCH

13.45 – 14.15Lessons from implementing Limbr, an interventionary research application”, James Kizer, Cornell TECH, New York

In conjunction with Prof. Deborah Estrin, et al., members of the Foundry at Cornell Tech implemented Limbr, an intervention for chronic pain. In this talk, I’ll lay out the system architecture and discuss our design decisions and the motivations behind them. I’ll highlight a few key insights after 6 months of operation, focusing on things we should have done differently and deployment issues we faced (running our Ohmage container in a HIPPA compliant cloud, onboarding patients, dealing with schema evolution). I’ll discuss the potential for ResearchKit / ResearchStack (RSuite) to enable the rapid prototyping of interventionary applications, as well as the needs of a backend to store, process, and share participant data while maintaining privacy and security. I will also mention new research directions we are exploring including the PePr privacy proxy (http://cornelltech.io/spice-up-your-personal-privacy-with-pepr/ ).

14.15– 14.45The Complexities of Mobile/Cloud Applications”, Håvard Johansen, Computer Science, UiT

Smart phones are great computational devices to utilize for health and wellness applications that must be highly available to their users. To help building such applications, large ecosystem of smart-phone systems, platforms, and services are available; each promising endless possibilities and easy development. In this talk I will describe some of the complexities of modern mobile/cloud application, using our current mobile/cloud athlete monitoring system as a use case. Although our app provides narrowly targeted functionality, it combines the Ohmage technology from Cornell TECH with a custom built hybrid mobile-phone front-end app and several additional back-end services. The talk will focus on the technology involved in running such apps, and the challenges we face.

14.45 – 15.15Use-based Privacy for Ohmage”, Eleanor Birrell, Computer Science, Cornell, Ithaca

Use-based privacy is the view that privacy is not just concerned with keeping certain information secret, but more generally it concerns restricting how information is used.  In order to enforce privacy, a privacy regime must enforce the appropriate use restrictions. In this talk, I will describe a policy language that supports an expressive class of use-based privacy policies, and I will describe how to extend the Ohmage-omh mobile health platform to enable use-based privacy. I will also discuss how trusted hardware—specifically the Intel Software Guard Extensions (SGX)—can be leveraged to enable data analysis and data sharing while enforcing the use restrictions imposed by use-based privacy policies.

15.15 – 15.45 Break (refreshments)

15.45 – 16.15     “Databox: Privacy-Aware Infrastructure for Managing Personal Data“, Hamed Haddadi, Queen Mary University of London

The Databox seeks to enable all the advantages of personal data analytics while at the same time enforcing accountability and control in order to protect a user’s privacy. In this work, we propose and delineate a personal networked device that allows users to collate, curate, and mediate their personal data. (http://www.databoxproject.uk/)

16.15 – 16.45Summary and Closing”; Bjørn Olstad, Microsoft Development Centre Norway; Dag Johansen, UiT

19.00 – 22.00                 Dinner Fjellheisen (http://fjellheisen.no/en/ )