Mri Image Reconstruction: Using Non Linear Conjugate Gradient Solution

The images extracted from Tomography are complex. Tomography images like Magnetic Resonance Images (MRI), angiograms, X-rays or CT scanned images extracted from scanner are further reconstructed to analyze smallest piece or portion of the images for investigation of diseases. The MR images when they extracted from scanner are under-sampled, added with artifacts called interference. L1-norm technique is used to minimize these interferences which are like random noise. A nonlinear conjugate gradient solution is adopted to minimize these noises.The natural images, they are real images, added with different noises like random / speckle / white noise /salt and pepper noise. If they are undersampled and incoherent are reconstructed by calculating the filter coefficients and denoising by conjugant gradient technique. A comparison of the PSNR, SSIM, and RMSE for natural (both color and gray) and medical images is tabulated. Matlab Algorithms are used in this work for reconstruction of images.