Lifespan AI - Our Mission:

The digital revolution is changing our world. Digital devices, sensors, and technologies casually acquire our data while evolving with us – so, digital natives leave a rich trace from the cradle to the grave. We believe that such collections of bio, social, and lifestyle signals will soon complement longitudinal studies, to improve healthcare through precise and personalised prediction, early prevention, and targeted intervention guided by causal inference. We are convinced that the massive influx of health data in combination with recent advancements in artificial intelligence (AI) holds great potential to transform the health sciences.In this research unit, we envision to advance AI methods and tools to model, predict, and explain health outcomes from multi-dimensional data that span the entire life. We aim to responsibly use unique data sources to extract relevant knowledge, to advance deep learning models, and to perform causal discovery with the ultimate goal to unravel the aetiology of complex diseases and to optimise prevention strategies.

Work Program:

The work programme consists of six projects that strive for our vision from different perspectives, organised in three themes: Data & Methods (D), Models & Interpretation (M), and Inference & Causality (C). Project D1 will advance deep learning strategies to explore and process long-term temporal changes based on the integration of multi-dimensional and multi-source data. Project D2 will combine neural networks and mixed effects models to predict individual health trajectories over the life-course. Project M1 will advance normalizing flow methods to derive joint distributions and conditional densities for health outcomes. Project M2 will create a cognitive digital twin from human everyday activities to predict changes across age groups and to simulate data for rare conditions. Project C1 will develop explainable AI methods for recurrent neural networks and time-to-event outcomes that are able to learn and adapt over time. Project C2 will establish a causal discovery framework for lifespan studies by combining datasets with age or time overlap and allowing for nonlinearities.The research unit draws expertise from applicants of both universities in Bremen and the Leibniz Institute for Prevention Research and Epidemiology – BIPS, who jointly cover disciplines ranging from mathematics and computer science to statistics and epidemiology. All applicants are located in Bremen and are already connected via institutional links, high-profile areas, joint centres, and various initiatives. This research unit is further strengthened by the newly established cooperation professorship on “Machine Learning in Statistics” at the interface between AI and epidemiology, which serves as the bridge professorship.

The application should include a CV, a publication list, and a cover letter that describes the motivation to join MMM and details how the applicant’s research agenda fits to the goals and mission of MMM. The acceptance of a new member is decided by the Executive Board with a two-thirds-majority requirement.