Fast Iteratively Reweighted Least Squares Algorithms for Analysis-Based Sparse Reconstruction

Ill-posed problems widely exist in medical imaging and computer vision. In order to seek a meaningful solution, regularization is often used if we have certain prior knowledge. With the emerging of compressive sensing (CS) (Candes et  al., 2006; Donoho, 2006), sparsity regularization has been an active topic in recent years. If the original data is sparse or compressible, it can be recovered precisely from a small number of measurements. The ℓ1 norm is usually used to induce sparsity and gains great success in many real appli cations.

Source: Medical Image Analysis - August 7, 2018 Category: Radiology Authors: Chen Chen, Lei He, Hongsheng Li, Junzhou Huang Source Type: research