Research

Causal AI and Clinical Decision-Making

I develop causal inference and AI methods for clinical and surgical applications.

Selected Work

• Causal Inference under Interference with Learned Exposure Mappings

• When Prediction Error Is Not Enough: Evaluating Nuisance-Function Prediction for Causal Estimation

• AI-Assisted Causal Inference and Mediation Analyses of Environmental and Psychosocial Determinants of Subjective Cognitive Difficulties in the All of Us Research Program

• Challenges and Recent Advances in Visual and Haptic Feedback for AI-Enhanced Teleoperated Robot-Assisted Surgery

Representation Learning and Multimodal Biomedical Data

Biomedical research increasingly brings together data from many sources, including clinical, laboratory, behavioral, socioeconomic, and genomic data.

I develop representation learning methods for integrating these data while capturing latent structure that can be useful for biomedical research. I am particularly interested in using learned representations to study complex relationships and potential mechanisms in high-dimensional health data.

Selected Work

• AI-driven Multimodal Representation Learning for Latent Mediation Structure Discovery of Socioeconomic Disadvantage, Psychosocial Factors, and Cardiometabolic Multimorbidity

• Different representation learning objectives recover distinct latent structures from the same psychometric data

Environmental and Population Health

Earlier in my research, I used machine learning and statistical methods to study environmental exposures, population health, and social vulnerability.

This work combined large-scale data on air pollution, weather, traffic, climate, health, and sociodemographic factors.

Selected Work

• Modeling impacts of traffic, air pollution, and weather conditions on cardiopulmonary disease mortality

• How to better predict the effect of urban traffic and weather on air pollution? Norwegian evidence from machine learning approaches

• Interaction between Climate Factors and Air Quality in Three Norwegian Cities: A machine learning analysis

• Physics-based machine learning for predicting urban air pollution using decadal time series data

• Integration of ten years of daily weather, traffic, and air pollution data from Norway’s six largest cities

• Inequitable efficiency: Unravelling the social and built environment drivers of London’s housing energy performance