US Workshop: Aiforia

Translate slides into discoveries & decisions

• Generating AI models in histopathology for research and clinical use is no longer only for dedicated large center teams
• We will show how you can get into quick, agile, and incremental development of AI models.
• Validation and continuous quality assurance is key for ongoing improvement and safe deployment. We will go through what, and how, you can do to ensure this.
• Are you looking to build a consortium to collaborate on deep learning projects? We will show you how easy it can be .

17:00 – 17:15 Opening remarks and intro to Aiforia
Thomas Westerling-Bui Director, Scientific Strategy and Business Development at Aiforia

17:15 – 17:45 Bridging the gap between digital pathology and molecular analysis
Quantumcyte has developed a unique platform that bridges the gap between digital pathology and molecular analysis. The Cytomask platform, developed by Quantumcyte, directly transfers the annotations generated by AI and digital pathology into molecular material (DNA, RNA, proteins) for further molecular analysis, thus retaining and spatial resolution and providing a high level of cellular purity. We demonstrate the Cytomask platform as integrated with Aiforia’s AI based annotations on cardiac disease tissue in collaboration with the University of Heidelberg.

John Butler, CSO & Bidhan Chaudhuri, CTO Quantumcyte

17:45 – 18:45 Developing AI-based quantitation of DSS colitis
We used Aiforia to develop an AI-based algorithm to analyze DSS colitis histopathology. The tool improves the efficiency of scoring DSS colitis and also screens a broad range of histopathological features to identify areas of interest for pathologists to perform manual follow-up. We will discuss the development, validation, and initial results generated by the algorithm.

Alex Klimowicz, Senior Principal Scientist, Boehringer Ingelheim Pharmaceuticals, Inc.

18:45 – 19:00 Break

19:00 – 20:00 Developing AI-driven quantitative histopathology in colorectal carcinoma
Colorectal carcinoma is histologically heterogeneous and current pathologic assessment does not capture this variability in a quantitative manner
Deep learning can provide a quantitative assessment of histologic features in colorectal carcinoma
Increased presence of poorly differentiated clusters/tumor budding, and immature stroma, were associated with reduced recurrence free survival

Rish Pai MD, PhD, Consultant Pathologist and Professor of Laboratory Medicine and Pathology at the Mayo Clinic

20:00 – 20:45 Agile AI model development for pre-clinical and primary diagnostics
Thomas Westerling-Bui Director, Scientific Strategy and Business Development at Aiforia

20:45 – 21:00 Close Out & QA with Aiforia