We are pleased to share that our researchers participated in ECDA 2026, presenting three contributions in the fields of privacy and security, statistical methodology, and clinical AI.
Prof. Dr. Kristina Yordanova presented “Watching the Watchers: Lessons Learned in Privacy, Security, Technology and Governance from Building and Operating Four Generations of Smart Labs”, together with Dr. Emma Tonkin. The work shares lessons learned from smart labs and sensor-equipped homes, focusing on privacy, security, ethics, governance, and informed consent.
Dr. Philipp Adämmer presented “StatsOP.jl: A Julia Package for (Sequential) Hypothesis Testing”, introducing a user-friendly Julia package for statistical inference in time series based on ordinal pattern analysis.
Mostafa Razavi Ghods presented “Generate Only What Matters: Comparing Token-level BIO and Span-JSON Generation on CARDIO:DE”. The work compares two approaches to clinical named entity recognition in German clinical notes. While token-level BIO currently achieves higher entity extraction accuracy, the proposed Span-JSON format produces 30.7% less target text, offering a more compact and potentially more efficient output for LLM-based generation.
We are proud to see our researchers contributing to discussions at the intersection of digital health, data science, statistical methods, and artificial intelligence.