Compressed sensing is a new paradigm that allows high resolution signals/images to be acquired using a very small number of measurements. This talk will present a brief overview of compressed sensing techniques, including their mathematical formulation and potential application in medical imaging. We will then discuss computational techniques for solving the difficult non-differentiable optimization problems that arise in this field. In particular, we shall focus on Split Bregman algorithms for L1 regularized problems.
Other applications of these numerical techniques include image segmentation, registration, denoising, and deblurring.
Back to Mathematical Problems, Models and Methods in Biomedical Imaging