ACM SIGKDD & Annual KDD Conference 2024 in Barcelona
The Department for Medical Informatics at the University Medical Center Göttingen recently delivered a lecture-style tutorial on Explainable AI (XAI) for Biosignals at the ACM SIGKDD & Annual KDD Conference 2024 in Barcelona. This tutorial was a collaborative effort between two of our leading research groups: the Biosignal Processing Group led by Dr. Nicolai Spicher and the Clinical Decision Support Group headed by Prof. Anne-Christin Hauschild. They were supported by Philip Zaschke, Miriam Cindy Maurer, and Philip Hempel on-side and Jacqueline Michelle Metsch off-side.
The tutorial attracted 33 participants from the broader AI and data mining communities, focusing on the critical role of explainability in AI models applied to biosignals, such as ECG and EEG. The key objective was to highlight how transparent and interpretable AI systems are essential for their successful integration into clinical practice. This is especially important in medical applications, where trust in AI-driven decisions can directly impact patient care.
This exchange between the fields of medical informatics and computer science is vital for both communities. While our expertise lies in the application of AI in the medical domain, collaborating with data scientists and AI researchers from diverse fields fosters innovation. For the AI and data mining community, understanding the unique challenges and high-risk environment of healthcare can lead to more robust and adaptable AI systems also in other application purposes. Conversely, the medical field benefits from the cutting-edge techniques and insights that computer science brings to the table.
For those who wish to delve deeper into the tutorial content, we have made the slides and Jupyter notebooks freely available online. Additionally, our comprehensive 7 pages survey paper “Explainable AI on Biosignals for Clinical Decision Support” is available via the conference proceedings.
Explore the tutorial resources: https://jacquelinebeinecke.github.io/xai-biosignal-cdss/
Read our survey paper: https://doi.org/10.1145/3637528.3671459
Through this tutorial and our ongoing research, we continue to bridge the gap between AI and healthcare, promoting an interdisciplinary approach that improves both fields and ultimately contributes to the development of AI systems that are effective, transparent, and trustworthy.
Links:
https://pad.gwdg.de/Ns1L9QB1QeavXE7TKv37tQ (for more photos)