Probabilistic Modeling of Anatomical Shape
Tom Fletcher, assistant professor, School of Computing and the Scientific Computing and Imaging Institute
University of Utah
Monday, May 12, 2014
Brickyard (BYENG) 420 [map], Tempe campus
Statistical models of shape have important applications in biology, medicine and computer vision. Recent work has focused on manifold representations of shape spaces, which can capture complex, nonlinear deformations between shapes and trajectories of shape change over time. Much of the current work frames the problem of model fitting as a geometric optimization, e.g., minimizing squared residuals to the data.
In this talk, Tom Fletcher will discuss recent developments in putting these models into a coherent probabilistic formulation. This includes Bayesian interpretations of shape models and inference procedures such as Markov Chain Monte Carlo for shape data. He will discuss several models, including simple averages (e.g., image atlases), factor analysis, and longitudinal models of shape evolution.
Tom Fletcher received his B.A. degree in mathematics at the University of Virginia in 1999. He received an M.S. in computer science in 2002 followed by a Ph.D. in computer science in 2004 from the University of North Carolina at Chapel Hill. He is currently an assistant professor in the School of Computing and the Scientific Computing and Imaging Institute at the University of Utah. His research is focused on creating novel methods at the intersection of statistics, mathematics, and computer science to solve problems in medical image analysis.
Hosted by Professor Yalin Wang