9.00 – 9.30 Brandon gallas - Evaluating Computational Pathology at the US FDA and Related Research
The FDA has been evaluating and regulating computer algorithms that process medical images for many years in radiology. While there are some algorithms that have been cleared or approved in digital pathology, the pace of submissions is expected to increase as the first whole slide imaging (WSI) system was permitted to be marketed in the US. In this talk, I will outline the basic algorithm types (computer-aided detection, CADe; computer-aided diagnosis, CADx; automated detection or diagnosis; semi- and fully-automated measurement or quantitation). I will also discuss how these algorithms fit into the US FDA regulatory framework. The pathways and methods used for evaluating algorithms in radiology guide us in developing the pathways and methods in pathology. I will provide a case study of one computer-aided detection (CADe) algorithm and outline its evolution over two decades. This evolution generalized the indications for use to new imaging systems (film to digital to 3D; one imaging system to many others), new performance claims, and new operating characteristics. I will also outline some regulatory and research programs and projects happening at the FDA related to computational pathology.
CV Brandon D. Gallas provides mathematical, statistical, and modeling expertise to the evaluation of medical imaging devices at the FDA. His main areas of research are in the design and statistical analysis of reader studies (https://github.com/DIDSR/iMRMC/releases, https://cran.r-project.org/web/packages/iMRMC/index.html), image quality, computer-aided diagnosis, and imaging physics. He began working on whole slide imaging devices as part of his regulatory responsibilities at the FDA in 2009. He turned his research effort in this direction and has been investigating the design, execution, and analysis of feature studies to compare pathologist performance and reproducibility using whole slide imaging devices and using the microscope. These studies are enabled by eeDAP, an evaluation environment that registers the digital images to the glass slides (https://github.com/DIDSR/eeDAP/releases). Dr. Gallas is organizer of a working group focused on the evaluation of whole slide imaging technical and pathologist performance (https://nciphub.org/groups/wsi_working_group). Before working at the FDA, he was in Dr. Harrison Barrett’s radiology research group at the University of Arizona earning his PhD degree in applied mathematics from the Graduate Interdisciplinary Program.