Compressively sampled MR image reconstruction using generalized thresholding iterative algorithm.

Area of Science:

• Medical Imaging
• Signal Processing
• Computational Science

Background:

• Compressed sensing (CS) is revolutionizing Magnetic Resonance Imaging (MRI) by enabling image reconstruction from limited k-space data.
• This reduction in data acquisition significantly shortens MRI scan times.
• Effective signal recovery in CS-MRI relies on sophisticated non-linear reconstruction algorithms.

Purpose of the Study:

• To introduce an improved iterative algorithm for CS-MRI image reconstruction using a p-thresholding technique.
• To enhance image sparsity, a crucial factor for successful CS reconstruction.
• To address the challenge of accurately reconstructing MR images from under-sampled k-space data.

Main Methods:

• Developed a novel iterative algorithm based on the p-thresholding technique, modifying the iterative soft thresholding algorithm (ISTA).
• The proposed algorithm minimizes non-convex functions to promote image sparsity.
• Validated the algorithm's performance using simulated and real MRI data.

Main Results:

• The p-thresholding iterative algorithm effectively reconstructs fully sampled MR images from under-sampled data.
• Quantitative analysis using Peak Signal-to-Noise Ratio (PSNR), Artifact Power (AP), and Structural Similarity Index Measure (SSIM) demonstrated superior performance.
• The proposed method outperformed existing iterative algorithms like log, soft, and hard thresholding techniques.

Conclusions:

• The p-thresholding iterative algorithm offers a significant advancement in CS-MRI image reconstruction.
• This method provides improved image quality and artifact reduction compared to conventional techniques.
• The algorithm's effectiveness is confirmed on both simulated and clinical MRI datasets.