February 11, 2009 Joint special session of the Conference on Computer-aided Diagnosis and Conference on Image Perception, Observer Performance, and Technology Assessment Session chair: Kyle J. Myers During more than forty years of professional life, Dr. Robert F. Wagner made numerous contributions to the field of medical imaging that significantly impacted academia, industry, and the regulatory science. Bob’s early work was devoted to the application of statistical decision theory to the evaluation of images degraded by measurement noise. He considered many modalities, tasks, and both human and machine observers, building broad consensus on a new, rigorous approach to the definition of image quality in medical imaging. Next, Bob and his colleagues took up the problem of patient variability, an additional source of “noise” in medical images. He then considered how to apply this same decision theoretic approach to the assessment of imaging systems confounded by artifacts. Most recently he was concerned with reader variability - where the readers could be doctors with a range of skills, or a set of computer algorithms trained with limited data. His unified approach created order out of the many problems of increasing complexity he tackled in his brilliant and memorable career. Bob’s first decade: in the beginning, Dave Brown, Ctr. for Devices and Radiological Health slides paper  Statistical ultrasonics: the influence of Robert F. Wagner, Michael F. Insana, Univ. of Illinois at Urbana-Champaign slides paper  NEQ: its progenitors and progeny, Harrison H. Barrett, College of Optical Sciences/The Univ. of Arizona slides paper  Performance-based assessment of reconstructed images, Kenneth M. Hanson, Los Alamos National Laboratory slides paper  Finding order in complexity, Kyle J. Myers, U.S. Food and Drug Administration slides paper  All papers referenced here were published in volume 7263 of the Proceedings of the SPIE. |