Comparison of Algorithms for Compressed Sensing of Magnetic Resonance Images

TL;DR

This paper compares three optimization algorithms for compressed sensing in MRI, demonstrating how they can reduce acquisition time while maintaining image quality, thus improving patient comfort and efficiency.

Contribution

It provides a comparative analysis of three popular optimization algorithms for MRI compressed sensing, highlighting their effectiveness in image reconstruction.

Findings

All algorithms successfully reconstructed images from fewer samples.

One algorithm showed superior reconstruction quality in most cases.

Results indicate potential for faster MRI scans with maintained image fidelity.

Abstract

Magnetic resonance imaging (MRI) is an essential medical tool with inherently slow data acquisition process. Slow acquisition process requires patient to be long time exposed to scanning apparatus. In recent years significant efforts are made towards the applying Compressive Sensing technique to the acquisition process of MRI and biomedical images. Compressive Sensing is an emerging theory in signal processing. It aims to reduce the amount of acquired data required for successful signal reconstruction. Reducing the amount of acquired image coefficients leads to lower acquisition time, i.e. time of exposition to the MRI apparatus. Using optimization algorithms, satisfactory image quality can be obtained from the small set of acquired samples. A number of optimization algorithms for the reconstruction of the biomedical images is proposed in the literature. In this paper, three commonly used optimization algorithms are compared and results are presented on the several MRI images.