The Diagnexia Computational Pathology Symposium is a pioneering one-day event bringing together experts in AI, digital pathology, and image analysis. This symposium will explore the latest advancements in computational pathology, including deep learning, data-driven diagnostics, and regulatory considerations.

TOPICS COVERED

• The Diagnexia Computational Pathology Symposium at Reuben College, Oxford, is a full-day event exploring AI and digital innovations in pathology. The day begins with introductory sessions on computational pathology, covering digital imaging, whole slide imaging (WSI), and image processing techniques such as segmentation and feature extraction.
• The morning continues with deep learning fundamentals, introducing neural networks, convolutional architectures (ConvNets, Transformers, ViTs), and performance evaluation. After lunch, discussions shift to data preparation, annotation challenges, and model training approaches, followed by validation and regulatory considerations for AI-driven pathology tools.
• Later sessions highlight pathology’s data-heavy nature, emphasizing emerging trends in AI, multimodal models, and the evolving role of pathologists. Real-world case studies demonstrate AI’s impact on diagnostics, leading into ethical considerations such as patient privacy and AI accountability.
• The day concludes with an open discussion and Q&A, fostering collaboration to advance computational pathology.

WHO SHOULD ATTEND?

Designed for pathologists, researchers, and data scientists, the event features expert-led lectures, interactive discussions, and real-world case studies. Attendees will gain valuable insights into how AI and digital tools are transforming pathology, improving diagnostic accuracy, and shaping the future of clinical practice. Join us to connect, collaborate, and advance the field of computational pathology.

SPEAKERS

• Pierre Moulin
• Kurt Zatloukal
• Runjan Chetty
• Donal O'Shea
• Mark Gregson
• John Weldon
• Emre Kose
• Adam Power

For more information, please contact: [email protected].