May 31, 2022, 14:00 CET

Data Privacy in Systems Medicine using Federated Learning

Prof. Dr. Jan Baumbach

Chair of Computational

Systems Biology Chair

University of Hamburg

Online Seminar via Zoom

Link and password upon request!

Please write an e-mail to: iem-seminar.umweltmedizin@tum.de

Data Privacy in Systems Medicine using Federated LearningArtificial intelligence offers game-changing opportunities to healthcare.However, it also harbors risks to patient privacy in particular when dealing withsensitive clinical data stored in critical healthcare IT infrastructure. Specifically,data exchange over the internet is perceived insurmountable, posing aroadblock hampering big -data-based medical innovations. We created a novelAIplatform, the FeatureCloud AI app store that is based on the idea offederated learning where only model parameters are communicated. Tomaximize privacy, sensitive datasets remain stored locally and are analysedbehind safe firewalls. We will exemplary investigate the power of FeatureCloudapps for decentralized (1) genome-wide associationstudies (GWAS), (2) geneexpression data mining, and (3) time-to-event data analytics to demonstratehow FeatureCloud may enhance worldwide collaboration, accelerateinnovation, and democratize scientific data usage.