MICCAI Fellows Talk with Stephen Pizer
Monday 1st December 2025
The MICCAI Fellows Talk Series highlights the work of our accomplished MICCAI Fellows. MICCAI Society Fellowships are awarded to a small number of senior members of the MICCAI community in recognition of their substantial scientific contributions and service to the MICCAI community.
We're excited to invite you to our next MICCAI Fellows Talk with:
Stephen Pizer, MICCAI Fellow 2010, Kenan Professor of Computer Science, University of North Carolina at Chapel Hill
Topic: An Object Representation for Statistics on Anatomic Objects
Monday, December 15, 2025 at 9:00 am EST / 3:00 pm CET
Abstract:
Abstract: A novel representation called the evolutionary s-rep represents anatomic objects by encompassing an unusually rich collection of geometric features, image intensity based features, genetic features, etc. It can be used for shape-based classification of anatomic objects, hypothesis testing on inter-class shape variations, and object segmentation. The evolutionary s-rep, example applications, and performance evaluations in comparison to other shape representations will be discussed.
This new representation provides better statistics than alternatives because its across-object positional correspondences are based on geometry not just on the object boundary but also in its interior, on second-order and not just positional (0th order) geometric features, and on finding for each case a diffeomorphism from a basic, ellipsoidal object. Also, a variant of the s-rep has been shown to provide object means that, unlike any other presently known shape representation, guarantee no local self-intersections. The software for fitting the evolutionary s-rep to a boundary mesh as part of the web-resident shape analysis toolkit at salt.slicer.org will be presented.
Biography:
Stephen Pizer is the Kenan Professor of Computer Science, University of North Carolina at Chapel Hill. He was awarded a MICCAI Fellow in 2010 for lifelong commitment to advance medical image computing, particularly methodologies for shape statistics and image analysis, and efforts to transfer image technology to the clinic.
His PhD dissertation, in 1967, was the first in medical image computing, and he has been academically working in that area continuously since his first job at Massachusetts General Hospital in 1962. In the UNC Computer Science faculty in 1967, he has had many collaborations across the UNC Medical School, most especially in neuroscience, radiation oncology, and most recently colonoscopy, and on software with Kitware. He has been principal advisor of 61 PhDs. In collaboration with Russell Taylor and Nicholas Ayache, he developed the plan for MICCAI conferences.
Prof. Pizer's research focused first on image quality restoration, then on 2D and 3D display, and also models of human vision. With Charles Metz, his 1971 writing provided a component of the EM algorithm. Since about 1980, his focus has been on geometric models of anatomic objects especially suited for statistics, their statistical analyses on non-Euclidean manifolds, and clinical applications. The s-rep based statistics on s-reps and object-relative intensities led to the commercial success of the spinoff, Morphormics, now part of Accuray.