Sparse Sampling MRI Reconstruction Technique via ANN

Abstract MRI is a viable tissue of an image. But in slow processing of image acquisition, MRI creates some noisy data. Compressed Sensing is a join mechanism of compression and sensing. So Compressed Sensing method is guiding to decrease the pattern recognition and suppress the some artifacts in MRI. However, MRI has some sparsity image which is obtained by the curvelet transformation and the reconstructed the noise data. But the issue of curvelet transformation is that it could not manipulate the sparse image. So the Patch Based Directional Curvelet Transformation (PBDCT) recovered the problem of that transformation in this research. But sparse sampling is built a hardware which is very difficult. Hence the proposed sparse sampling process is utilized for Artificial Neural Network (ANN) Experimentally, this proposed method is compared with SIDCT and PBDCT based on Signal-to-Noise Ratio (SNR) and Relative L2 Norm Error (RLNE)