Model Error in CT Reconstruction with Uncertain View Angles

Explore the challenges and solutions in computed tomography (CT) reconstruction with uncertain view angles in this one-hour virtual seminar. Delve into the world of inverse problems and their inherent instabilities, focusing on the often-overlooked impact of model errors. Learn about new methods for handling model errors in inverse problems, with a particular emphasis on CT imaging with uncertain measurement geometry. Discover how quantifying view angle uncertainty through a model-discrepancy formulation can significantly improve image reconstruction quality. Examine numerical experiments demonstrating enhanced relative reconstruction error and visual quality, and see how view angle estimation techniques can achieve results comparable to those obtained with correct view angles. Gain valuable insights into the importance of addressing model errors in inverse problems and their practical applications in medical imaging.

28th Imaging & Inverse Problems (IMAGINE) OneWorld SIAM-IS Virtual Seminar Series Talk