PhD Position in Machine Learning Preterm Lung Health

Department of Biomedical Engineering (DBE), University of Basel

Group:Analytics & Informatics for Child Health (AICH) in collaboration with University Children's Hospital Basel (UKBB)

Starting date:November 2026 or upon agreement

Duration:4 years

The Analytics & Informatics for Child Health (AICH) group develops AI/ML methods, digital tools, and secure data pipelines to advance pediatric healthcare. We work at the intersection of clinical medicine, machine learning, and data-intensive research, collaborating closely with hospitals, clinical IT teams, and research partners at the University of Basel.

Preterm birth affects around 14 million infants each year and is a major risk factor for lifelong respiratory disease. Current definitions of bronchopulmonary dysplasia (BPD) are categorical and predict long-term outcomes poorly. Robust early-life markers to identify infants at the highest risk of prematurity-associated lung disease (PLD) are lacking.

This project uses the Basel-Bern Infant Lung Development (BILD) cohort, which combines infant lung function, breath metabolomics, urine oxidative stress markers, clinical history, high-resolution air pollution exposure, and longitudinal lung function follow-up into adolescence. The goal is a continuous, physiology-based risk stratification framework to replace categorical BPD definitions, and to inform a respiratory follow-up program for preterm children in Switzerland.

The PhD student will lead the machine learning component: integrating clinical, physiological, environmental, and biomarker data to identify early-life signatures, aetiotypes, and treatable traits of PLD, with external validation in cohorts from Italy and Australia.

• Develop and apply multimodal ML methods to integrate clinical, breath metabolomics, biomarker, lung function, and environmental exposure data.
• Model lung function trajectories from infancy to adolescence, with a focus on moderate-to-late preterm children.
• Perform external validation with different cohorts and assess generalisability.
• Publish in ML and clinical/respiratory venues; present at international conferences.
• Work closely with clinicians, epidemiologists, and exposure scientists; contribute to reproducible, well-documented code.

• MSc in Computer Science, Data Science, Statistics, Biomedical Engineering, or a related field.
• Solid programming skills in Python; fluency with ML frameworks (PyTorch, scikit-learn) and statistical modelling.
• You are curious about how early life shapes lifelong health, and motivated to build ML that clinicians will actually use.
• Experience with tabular or longitudinal data, missing data, and model validation. Experience with small-sample or clinical/epidemiological data is a strong plus.
• Interest in interpretable ML, clustering/phenotyping, or time-series/trajectory modelling.
• Interest in clinical research and respiratory or pediatric health; no medical background required.
• Rigour, independence, and the ability to communicate with non-technical collaborators.
• Fluent in English (spoken and written). German is an advantage but not required.

• A PhD at the interface of machine learning, clinical medicine, and environmental health, with direct translational impact.
• Access to a unique longitudinal birth cohort spanning infancy to adolescence, plus international validation cohorts.
• Close supervision and collaboration with clinical and technical experts.
• Access to state-of-the-art HPC infrastructure.
• Enrolment in the Graduate School Medical and Health Sciences (GMHS) , with training and conference opportunities.
• Supportive team culture and employment conditions in accordance with the University of Basel.