On digital and computational pathology - A paradigm shift towards pathology captured in pixels

B. Sturm

• Promotor: J. van der Laak
• Copromotor: W. Blokx
• Graduation year: 2025
• Radboud University, Nijmegen

Abstract

In chapter two and three, validation studies concerning the concordance in assessment of glass and digital slides of respectively biopsies from the larynx and colon, are described. In chapter four, a topic of concern in the pathology field that is the loss of multi-level focusing in regular WSIs is addressed. We explore the diagnostic value of z-stack scanning concerning diagnostic accuracy in melanocytic lesions of the skin. In chapter five, the potential added value of computer assisted mitosis identification in melanocytic lesions of the skin is assessed, based on an artificial convolutional neuronal network algorithm. In chapter six, we explore the concept of artificial intelligence prediction of neo-adjuvant chemotherapy outcome in triple negative breast carcinoma. Chapter seven contains a general discussion of the main findings of this thesis.