Abstract
At this moment, databanks worldwide contain brain images of previouslyunimaginable numbers. Combined with developments in data science, these massivedata provide the potential to better understand the genetic underpinnings ofbrain diseases. However, different datasets, which are stored at differentinstitutions, cannot always be shared directly due to privacy and legalconcerns, thus limiting the full exploitation of big data in the study of braindisorders. Here we propose a federated learning framework for securelyaccessing and meta-analyzing any biomedical data without sharing individualinformation. We illustrate our framework by investigating brain structuralrelationships across diseases and clinical cohorts. The framework is firsttested on synthetic data and then applied to multi-centric, multi-databasestudies including ADNI, PPMI, MIRIAD and UK Biobank, showing the potential ofthe approach for further applications in distributed analysis of multi-centriccohorts