Marie Piraud

AI Consultant Team Leader at Helmholtz AI
Address: Helmholtz AI Central Unit, Ingolstädter Landstraße 1, D-85764 Oberschleißheim, Germany

& Guest researcher at Technische Universität München
Chair for Computer Aided Medical Procedures & Augmented Reality


I joined the new adventure of Helmholtz AI in September 2020 and started leading the consultant team for Health. I am looking forward to new exciting challenges.


Research interests

I am a theoretical physicist who enjoys working at the frontier with other research fields. I am fascinated by complex and multi-scale systems, which designate systems whose modeling remain a challenge even when we understand the behaviour of all their small scale components. In most biological and biomedical problems, the descriptions at different scales (e.g. microscopic and macroscopic) are indeed difficult to relate to each other. And it is also the case of most condensed-matter systems, which are made of many particles whose individual properties are well characterized, but for which the description of the collective behaviour is often a challenge. Those are difficult topics in general for which both analytical and numerical modeling are as important as experimental observations.

I played with those ideas in the quantum realm, as I worked at the frontier between the fields of cold atoms and condensed matter for seven years. I first studied the role of microscopic disorder in quantum systems, which can have dramatic and non-intuitive consequences that do not exist in the classical world: it can transform a metal into an insulator, a phenomenon called 'Anderson localization'. I also studied the role of topology in quantum systems, as I investigated complex interacting sytems in the presence of gauge fields, which are known to induce global topological transitions. Both thematics are particularly relevant in the solid state, where disorder and magnetic fields are ubiquitous, and are now studied in ultra-cold atom experiments. In this context I derived analytical models and carried-out large-scale numerical simulations, which forged my taste for conducting those efforts in parallel.

I am now exporting those ideas and competences to the medical realm, in which systems are also studied at different scales. For example bridging the gap between small-scale biophysical models and medical observations at the scale of an individual or a population can prove difficult. To this aim, I am using the traditional framework of statistical inference as well as developping new methods combining model-based approaches and data-based machine learning techniques. Indeed, joining forces of the previous knowledge acquired on the underlying processes with powerful techniques of artificial intelligence is, in my opinion, a very exciting and meaningful avenue to better predictions.

Updated 2021