A new 2-D convex combination of recursive inverse algorithms

De-noising magnetic resonance images (MRI) has recently become an interesting topic in medical diagnosis applications. Many algorithms have been proposed for this purpose. However, these algorithms usually suffer from poor performance or time consumption. In this paper, we propose a 2-D version of the recently proposed convex recursive inverse (RI) algorithm that provides fast convergence at the beginning to save time and then provides high performance in terms of noise removal. To test the algorithm, a de-noising experiment has been conducted on MR image that is assumed to be corrupted by an additive white Gaussian noise (AWGN). Simulations show that the proposed algorithm successfully recovers the image.