Research

We develop mechanistic models to understand pathophysiological mechanisms, focusing on dynamic immune responses to external stimuli (pathogens, environmental factors, treatments).

We apply statistical modelling and machine learning techniques to analyse data, such as: Time-series forecasting, to predict the evolution of diseases like atopic dermatitis. Bayesian modelling, to learn from small data, quantify uncertainty and develop flexible models that reflect data-generating mechanisms.

We develop computer vision pipelines to support medical/experimental data collection, such as EczemaNet (automatic assessment of eczema severity scores from camera images) and AFNet (automatic quantification of A. fumigatus fungal burden from histology images).