Overcoming Real-World Variability: Inside Pathology's Most Comprehensive AI Validation Study

Recorded live on March 31, 2020


The accelerated pace of artificial intelligence development in pathology has raised serious questions about the performance of computational applications in practice. Algorithmic sensitivity and specificity, alongside transferability and the capacity to address variation in biopsy methods, tissue preparation and staining, and digital scanning processes, have critical implications on the viability of AI in real-world laboratories.  

In the most comprehensive AI validation study conducted in pathology to date, researchers at Proscia and collaborators in clinical practice demonstrated the ability of AI to address these sources of variability while achieving up to 98% classification accuracy across laboratory settings. Join Proscia’s Chief Medical Officer, Mike Bonham, MD, Ph.D., and Director of AI Research, Rajath Soans, Ph.D., for a webinar as they explore:

  • Considerations that influenced the design of Proscia’s seminal study
  • Key findings from the Scientific Reports publication 
  • How this study provides a framework for demonstrating the generalizability of computational applications in pathology


Watch Now

Submit the form below to view the previously recorded webinar.