Energy preserved sampling for compressed sensing MRI.

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Signal Processing

Background:

• Compressed Sensing Magnetic Resonance Imaging (CS-MRI) is crucial for faster image acquisition.
• Traditional random sampling patterns in CS-MRI lead to suboptimal image reconstruction due to neglecting k-space energy distribution.
• Variable density (VD) sampling patterns were developed to address this, but further optimization is needed.

Purpose of the Study:

• To introduce a novel energy preserving sampling (ePRESS) method for CS-MRI.
• To improve the cost function for CS-MRI reconstruction by incorporating phase correction and a region of support matrix.
• To develop an iterative thresholding algorithm (ITA) for efficient reconstruction using the improved cost function.

Main Methods:

• Proposed the energy preserving sampling (ePRESS) method for k-space data acquisition.
• Enhanced the cost function with phase correction and region of support matrix.
• Developed and applied an iterative thresholding algorithm (ITA) for image reconstruction.
• Evaluated performance using 2D digital phantoms and 2D in vivo human brain MR images.

Main Results:

• The ePRESS sampling method demonstrated superior performance compared to VD, POWER, and BKO sampling patterns.
• The improved cost function yielded higher quality MR image reconstructions than the conventional approach.
• The ITA reconstruction algorithm exhibited faster computation times than SISTA and was competitive with FISTA.

Conclusions:

• The proposed ePRESS sampling, improved cost function, and ITA reconstruction algorithm collectively enhance CS-MRI performance.
• This integrated approach offers a promising direction for optimizing CS-MRI acquisition and reconstruction.
• The findings suggest significant improvements in both image quality and computational efficiency for CS-MRI.