Sparse BLIP: BLind Iterative Parallel imaging reconstruction using compressed sensing.

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Imaging Reconstruction
• Signal Processing

Background:

• Parallel imaging accelerates MRI acquisition by using multiple receiver coils.
• Accurate coil sensitivity estimation is crucial for minimizing artifacts in parallel imaging reconstruction.
• Existing methods often struggle with precise sensitivity estimation, leading to image distortions.

Purpose of the Study:

• To develop a novel sensitivity-based parallel imaging reconstruction method.
• To simultaneously reconstruct coil sensitivities and MR images iteratively.
• To leverage prior information for enhanced reconstruction accuracy.

Main Methods:

• Proposed a new algorithm: Sparse Blind Iterative Parallel (SBIP).
• SBIP performs blind iterative parallel imaging reconstruction using compressed sensing principles.
• Reconstructs sensitivities and images simultaneously from undersampled data, enforcing sparsity on the image and constraints on sensitivities.

Main Results:

• SBIP demonstrated improved reconstruction performance compared to Sparse SENSE, JSENSE, IRGN-TV, and L1-SPIRiT.
• The proposed method achieved better results with the same number of measurements.
• Phantom and in vivo experiments validated the effectiveness of SBIP.

Conclusions:

• The Sparse Blind Iterative Parallel algorithm effectively reduces reconstruction errors.
• SBIP outperforms current state-of-the-art parallel imaging reconstruction methods.
• This method offers a promising approach for more accurate and artifact-free parallel MRI.