Data loss in biomedical and clinical research occurs primarily through scattering across disconnected systems and inadequate capture methods. Hospitals and clinics are now making efforts to centralise their data storage, but this fails to resolve the underlying problem: researchers and clinicians work predominantly with non-normalised, tabular, two-dimensional data structures that cannot represent the hierarchical nature of biological and clinical information. The consequences are measurable. Clinical datasets contain insufficient variables and systematic under-representation of several intersectional variables regarding patients, e.g. gender, or environmental conditions. These datasets lack harmonisation, preventing meaningful cross-study comparisons and meta-analyses. In this workshop, we will introduce the current state of clinical data repositories and explain how we can help non-technical users from the health sector become acquainted with accessible structured data formats. We show how moving from spreadsheets to structured formats helps create FAIR (Findable, Accessible, Interoperable, Reusable) data. We will discuss what currently exists to bridge the gap between standard protocols and the new demands of AI-powered research and data analysis, and brainstorm what the community could do to bring the medical and IT communities closer together.
Zugang zum Zoom-Raum und weitere Informationen zur Love Data Week 2026 gibt es hier.