Compressed Sense in Magnetic Resonance Imaging

Abstract

This work intends to illustrate the technique of Compressed Sense (CS), an innovative method introduced in the field of signal processing, which allows you to capture signals and images with far fewer samplings than “needed”, reducing scan times up to 50% and offering accurate and high-resolution images.

The Compressed Sense technique can be successfully applied for Imaging in Magnetic Resonance as it satisfies the properties of sparsity and incoherence. One of the primary requirements of the CS is based on the use of particular mathematical bases, whose function will be to represent the sampled functions. Between these, the Wavelet bases are of considerable importance as they subdivide the data in question into components of different frequencies, and therefore allow for the study of every element in the frequency spectrum with a resolution matched to its scale, such as, for example, those that are used in the encoding of images with the JPEG-2000. The Compressed Sense technique has allowed us to overcome the intrinsic physical limits inherent in the matter, developing a technique that continues to evolve and expand, improving and refining.