09.00 - 09.15 Jeroen van der laak - The TIGER challenge: automated quantification of TILs in breast cancer H&E images
Much progress in the field of Computational Pathology may be attributed to so-called ‘grand challenges’, international competitions that invite researchers in AI to solve clinically relevant problems. These challenges are very attractive for the AI community, because they present a well-defined and relevant problem, and give access to the data required to build a solution as well as give a framework to evaluate the AI performance. In this three part session, of which the final part will be presented by Prof Jeroen van der Laak, the concepts of grand challenges are explained, examples of previous challenges are discussed, and the results are shown of two very recent challenges: CONIC and TIGER.
Jeroen van der Laak is Principle Investigator and Professor of Computational Pathology at the department of Pathology of the Radboud University Medical Center in Nijmegen, The Netherlands and guest professor at the Center for Medical Image Science and Visualization (CMIV) in Linköping, Sweden. Prof van der Laak holds an MSc in computer science and acquired his PhD from the Radboud University in Nijmegen.
His research focuses on the use of machine learning for the analysis of digitized microscopic tissue sections (whole slide images). His research group was among the first to show the large potential of deep learning algorithms for analysis of whole slide images. Further research focused on improvements in deep learning strategies to increase robustness and accuracy, as well as on application of deep learning for various tasks in histopathology. Application areas include: improvement of routine pathology diagnostics, objective quantification of immunohistochemical markers, and study of novel imaging biomarkers for prognostics. He successfully developed models for the analysis of breast and colorectal cancer and for renal transplant biopsies. In 2016 and 2017, he coordinated the highly successful CAMELYON grand challenges. These worldwide competitions resulted in the first publication showing that deep learning can perform on par with pathologists for specific tasks in histopathology. The CAMELYON data set is among the largest and most studied in computational Pathology.
Prof van der Laak co-authored 100 peer-reviewed publications and is member of the editorial boards of Laboratory Investigation and the Journal of Pathology Informatics. He is member of the board of directors of the Digital Pathology Association and organizer of sessions at the European Congress of Pathology and the Pathology Visions conference. Dr van der Laak acquired research grants from the European Union, the Dutch government and the Dutch Cancer Society, among others. He collaborates in public-private partnerships, aiming to productize developed algorithms. Dr van der Laak is frequently invited as a speaker at international conferences.