
Time: 17:00
Location: T S Eliot Lecture Theatre
Speaker: Professor Mireia Crispin-Ortuzar
Oxford DPhil in Experimental Particle Physics (2015)
Assistant Professor and Group Leader, Early Cancer Institute, Department of Oncology, University of Cambridge
Co-Lead, Ovarian Cancer Programme & Mark Foundation Institute, Cancer Research UK Cambridge Centre
Chief Digital Officer, 52North Health
When a pathologist looks down a microscope at a tumor sample, what do they actually see, and what decisions depend on that moment? This lecture takes the pathology slide as a starting point to explore how algorithms are increasingly used, and sometimes not used, in cancer care. AI for image analysis is everywhere in research and the media, but where does it genuinely appear in clinical practice, and where does it fall short? Using examples from computational pathology, medical imaging, and the integration of molecular and clinical data, I will discuss how algorithms attempt to extract information across spatial scales, from cells to whole tumors, and why this remains a hard problem. We will examine what is missing for AI to be reliably "always on call" in oncology, and why algorithms are more likely to change how clinicians work than to replace them.
View more information on The Ockham Lecture Series