Chris Hart is a Senior AI Research Scientist at Babylon Health. Chris was educated at Oxford (MPhys Physics) and University College London (MSc Computational Statistics and Machine Learning), and prior joining to Babylon three years ago worked in the insurance sector on longevity risk and software engineering applications. Chris’s research at Babylon encompasses methods for assessing and stratifying health risks in populations, models for predicting disease risks using causal inference, and advanced evaluation techniques for medical diagnostic systems.
Prediction of future event times for e.g. hospital admissions or a novel diagnosis have great importance in healthcare and are typically studied under frameworks such as survival analysis. These methods on their own lack the ability to give individualised measures of interventional impact of a given treatment, such as a new medication or lifestyle change. Advances in causal inference have been introduced to better model such impacts and quantify measures such as individual treatment effects when working with observational data. This talk will review such methods and their application in modelling and evaluation of survival analysis tasks.