The first international workshop on geometric deep learning in medical image analysis (GeoMedIA) was held as a full-day event at Hotel Casa in Amsterdam, the Netherlands on November 18, 2022. The workshop brought together over 210 attendees interested in the fields of geometric deep learning and medical image analysis. The workshop is a result of a collaboration between Erik Bekkers from University of Amsterdam (an ELLIS Unit Amsterdam Faculty), Angelica Aviles-Rivero from University of Cambridge, and Jelmer Wolterink from the University of Twente. The ELLIS Unit Amsterdam proudly sponsored the event by, among others, providing 27 travel grants through the ELLIS Unit Amsterdam’s Workshop Programme.
The workshop aimed to draw attention to current developments in geometric deep learning for medical image analysis. Its core objective is to inspire researchers and expose the vast richness of geometric structure in medical image data through a day of exciting keynotes, contributed talks, and a panel discussion with experts. Topics included but were not limited to group convolutions, mesh CNNs, or graph neural networks with geometric priors.
In total, 25 works were presented at the workshop: 14 full-length papers, and 11 extended abstracts. The full papers are to be published in the Proceedings of Machine Learning Research (PMLR) with volume number 194. The workshop acknowledged and awarded the best paper to Mohamed A Suliman from Imperial College London.
Two noteable speakers, Prof. Emma Robinson from the King’s College London, and Prof. Michael Bronstein from the University of Oxford delivered their keynote presentations at the workshop. The event was concluded with a panel discussion with participation from industry, healthcare, and academia, including:
- Clarisa Sanchez, Professor at UvA and AUMC for her expertise in AI for Medical Imaging (+experience with clinical partnerships);
- Taco Cohen, machine learning researcher / principal engineer at Qualcomm AI Research, for his expertise in geometric deep learning and company perspective;
- Emma Robinson, Professor at King’s College London, for her expertise in geometric deep learning in medical imaging (+experience with clinical partnerships);
- Gabriele Cesa, PhD student at UvA and Qualcomm, for his contributions in opensource libraries and as geometric deep learning researcher; and
- Jorrit Glastra, CTO at Quantib, for his experience in AI-based solutions in health care (medical image analysis and quantification) and perspective from the industrial point of view.
More information on the GeoMedia workshop 2022 can be found here.
The ELLIS Unit Amsterdam continues to support the organization of workshops on AI which can promote the advancement of AI, particularly in the AI ecosystem here in Amsterdam. Find out more about the ELLIS Unit Amsterdam’s Workshop Programme here.