The Modeling Lab of ELTE Faculty of Informatics at EMBC 2025
The research team—Gergő Bognár, Dominika Darabos, and Gergő Galiger—represented the group and shared results from ongoing projects focused on enhancing the reliability and interpretability of neural networks in clinical and experimental settings.
Dominika Darabos presented:
"Explainable Spike Detection Algorithm on C-fiber Microneurography Data using Weighted Hermite Variable Projection Neural Networks"
This work introduces a novel spike detection framework for peripheral nerve recordings by fusing the variable projections neural networks with knowledge-augmented learning.
We’re proud to share that Dominika Darabos was selected as a finalist in the EMBC Student Paper Competition, recognizing the quality and impact of her contribution. Congratulations!
Gergő Galiger showcased:
"Sparse Dictionary Learning Neural Networks for ECG Signal Denoising"
This research merges sparse signal modeling with neural networks via deep unfolding the well-known FISTA algorithm to develop efficient and generalizable ECG denoising methods.
These projects are the result of close and productive collaborations with several leading clinical and academic institutions:
- RWTH Aachen University Hospital
- University Hospital Cologne
- University Hospital Würzburg
- Department of Medical Informatics, University Medical Center Göttingen
Gergő Bognár co-chaired the minisymposium on the EU AI Act: Opportunities and Challenges for Researchers in Building Trustworthy AI, a timely and important discussion on aligning academic innovation with emerging regulatory frameworks.
He also co-moderated, together with Declan O'Loughlin, the Young Professionals Roundtable and Whitepaper Initiative on the evolving role of consumer electronics in healthcare.
The roundtable covered a wide spectrum of technologies—from smartwatches and rings to UWB-based monitoring and tracking—and addressed key challenges and opportunities:
- How consumer electronics can improve accessibility and continuous monitoring
- Safety and ethical concerns in unregulated environments
- Best practices for interpreting and reporting results from consumer-grade data
Participants engaged in a lively, forward-looking discussion on the benefits, risks, and future directions of integrating consumer devices into healthcare and clinical workflows.
We’re grateful for the opportunity to share our work and ideas with the global biomedical engineering community, and inspired by the conversations, collaborations, and insights from #EMBC2025.