Session: Artificial Intelligence: What you need to know and its impact on the future of pathology
In this poster, we discuss the process for introducing a “hybrid gross anatomy program” – one that combines virtual models (as in those generated from CT scans) with real, cadaveric models. Our goal is to combine the kinesthetic learning mode of physical dissection with the full understanding of the anatomy using computational tools.
In this talk, I discuss a project that utilizes both traditional and modern (deep) learning methods for identifying patients at risk for loco-regional recurrence for oral cavity cancers. I cast the problem of classifier training as one of pedagogy as well as engineering, and present the concept of an “AI School for Pathology”, where human teachers try to efficiently improve the performance of their neural network students.
I am Assistant Professor of Pathology and Anatomical Sciences, Biomedical Engineering, and Biomedical Informatics at the University at Buffalo, SUNY. My lab develops computational tools for medical data, with a focus on imaging, machine learning, and artificial intelligence.