On 3rd, June, Dr. Pei-Yun Sabrina Hsueh gave a talk titled ''Transforming big data to actionable small data: Opportunities and Challenges of putting patient-generated health data in action''.
Associate Prof. Bian Yang moderated the lecture. Here is the abstract of this lecture:
The rise of consumer health awareness and the recent advent of personal health management tools (including mobile and health wearable devices) have contributed to the recent shift transforming the healthcare landscape. Despite the rising trend of health consumers, the impact of user-generated health data remains to be validated. Past research has shown that 60% of health determinants are related to exogenous factors including social, environmental and behavior factors. However, how to measure and incorporate these factors into standard care practice is still an emerging field. In fact, many applications are hinged on the lack of a good platform that can provide better integration and sense-making of a multitude of “exogenous" patient-generated health data (PGHD) sources. In this talk we shall discuss both the success stories and the areas that fall short with the use of PGHD, walk through our experience with generating insights from PGHD, and discuss the associated challenges. In particular, we shall review the proposed behavioral learning and adaptation platform to make PGHD effective for the next-generation behavioral healthcare systems “on the cloud” as well as for the on-device intelligent systems “on the edges.” First, we evaluate the need of integrating multiple sources and the methodologies for understanding individual risk factors. Secondly, we optimize and adaptive deployment of behavioral interventions based on a more granular level of behavioral input and ecological momentary responses. Finally, we bridge the gap between system and user initiatives by creating systems that can think "outside the box" to find an alternative solution that would be more amenable to this individual by self-experimentation. The integrated platform enables the recommendation of care plans based on population data and adapts them plans with emerging evidence at the individual level. Finally, key dimensions such as including security and trust management issues will be discussed in the context of creating the behavioral learning and adaptation platform to make sense of PGHD.