EGENIS seminar: "AI in Medicine: Finding equilibrium in global data capture", Prof Robin Pierce (University of Exeter)

The range of applications of AI in medicine has grown considerably in recent years. The increase in computational capacity has allowed for an array of technologies that can uncover a vast number of correlations that could improve health outcomes or yield scientific knowledge. Increased understanding of the impacts of the social determinants of health, environmental, and other (non-) biological factors on health outcomes would seem to support the drive to amass, aggregate, and integrate different types of data. Yet, even in global data capture, what is absent may be the greater challenge for data governance, possibly affecting explainability, accuracy and, ultimately, health outcomes. Data governance aims to govern data but may have little to say about “absent” data. Using examples of data-intensive technologies, e.g., Deep and Frequent Phenotyping, this paper explores the terrain of finding equilibrium as a regulatory challenge for health research.