Application of Hybrid Genetic Algorithm for Successful CS-MRI Reconstruction

G. Shrividya and S.H. Bharathi

In Magnetic Resonance Imaging (MRI) major concern is the huge data to be handled and the imaging speed. For faster image acquisition and image reconstruction Compressive Sensing (CS) is used in MR imaging process. Compressive Sensing (CS) is widely used signal processing method that has widespread applications in various fields from communications to image processing. Major issue that obstruct the application of compressed sensing in the field of medical imaging is the reconstruction time and cost of the system. This work proposes a hybrid Particle Swarm Optimization and Grey Wolf Optimization (HPS-GWO) algorithm for there construction of MRI images. Hybrid Walsh Hadamard Transform and Discrete Wavelet Transform (HWHDWT) is used as the sparsifying transform. The experimental results are analysed in terms of PSNR, MSE, SSIM and reconstruction time for the compressive sensed MRI reconstruction. The qualitative and quantitative analysis is carried out on data set collected from a hospital uphold the novelty of the proposed algorithm.

Volume 12 | Issue 3

Pages: 408-414

DOI: 10.5373/JARDCS/V12I3/20201208