Compressed sensing MRI using Singular Value Decomposition based sparsity basis.

Area of Science:

• Medical Imaging
• Signal Processing
• Biomedical Engineering

Background:

• Magnetic Resonance Imaging (MRI) is crucial but limited by slow data acquisition.
• Compressed Sensing (CS) accelerates MRI by exploiting signal sparsity.
• Existing CS-MRI methods often use predefined sparsifying transforms, limiting their effectiveness.

Purpose of the Study:

• To introduce Singular Value Decomposition (SVD) as a data-adaptive sparsifying transform for Compressed Sensing MRI.
• To evaluate the performance of SVD against conventional predefined transforms in CS-MRI.
• To demonstrate SVD's potential to improve image quality and representation sparsity.

Main Methods:

• Proposed Singular Value Decomposition (SVD) as a data-adaptive sparsifying basis for CS-MRI.
• Compared SVD with commonly used predefined sparsifying transformations (e.g., DCT, DWT).
• Evaluated the methods based on the sparsity of MR image representations and resulting image quality.

Main Results:

• SVD demonstrated superior performance in achieving sparser representations for a wider range of MR images compared to predefined transforms.
• The proposed SVD method resulted in improved image quality in Compressed Sensing MRI.
• SVD proved to be a simple yet effective alternative for sparsifying transforms in CS-MRI.

Conclusions:

• Singular Value Decomposition is an effective data-adaptive method for Compressed Sensing MRI.
• SVD offers advantages over predefined transforms by providing better sparsity and image quality.
• This approach presents a promising alternative for accelerating MRI acquisition using Compressed Sensing.