Real-World Data and Federated Analytics
One key feature of our research is using data collected from the world around us. This could be de-identified patient records, weather station measurements or satellite imagery. This real-world international evidence uses federated analytics to safely and securely combine individual person-level data into valuable learning.
These days, so much of our lives produces data. At PHI, we use this routinely collected data and combine it with the latest analysis pathways from artificial intelligence and medical statistics. Using data collected down to the level of individuals, we can then produce useful resources like clinical models for predicting health outcomes.
To preserve an individual's privacy in this process, we use de-identified data and federated analytics. Federated analytics enables data analysis without centralising it. Instead of sending personal data to a central server, it remains on users' devices, and algorithms process the data locally. From this analysis, we can produce aggregate trends which can be used in research, but which are anonymous and in which no individual's data can be identified, making it a promising solution for data-driven applications.
Many of our research projects in this area also focus on COVID-19, as PHI continues to aid efforts to understand and combat the pandemic. See more about our pandemic response here.
Featured Projects
 |
|
 |
|
LINKS AND PUBLICATIONS FROM OTHER PROJECTS
Explore the PHI Lab
Latest from PHI - PHAM2024
PHI is at PHAM, presenting the Lab's research and discussing AI, human choice and planetary health. We've also had our malaria project abstract, which was accepted to the conference, published in the Lancet Planetary Health as part of PHAM. See more of what we're getting up to on our blog.