Enhanced MRI Reconstruction using Compressed Sensing and Sparse Representation Techniques

In this paper, we propose a novel method for image reconstruction in the context of magnetic resonance imaging (MRI). The goal of our work is to improve the reconstruction quality while reducing the number of measurements that is required, and thus enabling faster imaging thus reducing the burden on patients. An Edge Enhanced Sampling pattern is introduced in this paper,a custom pattern based on the image's edge map, which utilises the edge information and develops the sampling pattern. We compare the method proposed with three existing patterns like random, Cartesian grid and radial and demonstrate its better performance. These patterns determine which measurements are acquired during the MRI scan. We generate these patterns based on the desired acceleration factor, which represents the reduction in the number of measurements compared to the original Nyquist rate. Overall, our work demonstrates the effectiveness of the proposed method in achieving high-quality image reconstruction from undersampled k-space data in MRI. This has the potential to significantly enhance the efficiency and speed of MRI scans, leading to improved patient experiences and better healthcare outcomes.