Image Reconstruction

About

Invention Summary: Magnetic Resonance (MR) imaging has been widely used in medical diagnostics because of its non-invasive manner and excellent depiction of soft tissue changes. Image reconstruction methodologies are vital to many applications in medical imaging. Recent developments in compressive sensing theory show that it is possible to accurately reconstruct the Magnetic Resonance (MR) images from highly undersampled K-space data and therefore significantly reduce the scanning duration. Researchers at Rutgers University have developed a sparsity technique for efficient MR image segmentation and reconstruction. Sparsity is a way of looking at a small part of a video signal to develop a full image. It only works if it is the small, but significant part of the image. The method is akin to how human vision cannot process every detail but explores information in a small area of what the observer sees. The method mathematically allows reconstruction of a full image based on the sparse data set (ie: Reconstructing an MRI data set from sampled data along radial lines). The new MR image reconstruction method is based on the combination of variable and operator splitting techniques. The algorithm minimizes a linear combination of three terms corresponding to a least square data fitting, total variation (TV) and L1 norm regularization. This has been shown to be very powerful for the MR image reconstruction. First, the original problem is decomposed into L1 and TV norm regularization subproblems respectively. Then, these two subproblems are efficiently solved by existing techniques. Finally, the reconstructed image is obtained from the weighted average of solutions from two subproblems in an iterative framework.   Market Application: Medical MR Image reconstruction, Homeland Security, borders or battlefield surveillance   Advantages: This method is shown to impressively outperform the classic methods and also two of the fastest methods in terms of both accuracy and complexity.