The symposium focused on the latest breakthroughs in genomics, computational pathology and the use of electronic health records through the lens of foundation models. Co-Chair of the Bavarian AI Council Fabian Theis introduced the audience to the world of foundation models in the healthcare sector, highlighting not only the potential of these models, but also the challenges they present. For example, foundation models require extensive data and computing resources and ethical considerations must also be taken into account, particularly in connection with sensitive health data.
Tingying Peng presented HistoGPT for generating accurate pathology reports and emphasized the precision of the model and its potential to transform diagnostics. Paulina Szymczak from the University of Warsaw presented her research on generative modeling for the discovery of antimicrobial peptides and offered promising solutions to the growing problem of antibiotic resistance.
The poster session brought together researchers from different disciplines and promoted exchange on the further development of the next generation of Health Foundation Models.
