Compressed sensing plus motion (CS + M): A new perspective for improving undersampled MR image reconstruction.

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computational Imaging

Background:

• Undersampled Magnetic Resonance Imaging (MRI) data acquisition presents significant challenges for high-quality image reconstruction.
• Dynamic MRI scans are particularly susceptible to motion-induced blurring and loss of fine details, degrading diagnostic accuracy.

Purpose of the Study:

• To develop an advanced framework for dynamic MRI reconstruction that simultaneously addresses image quality and motion artifacts.
• To introduce a novel multi-task optimization model, Compressed Sensing Plus Motion (CS+M), for improved undersampled MRI data reconstruction.

Main Methods:

• Propose a unified optimization problem to concurrently reconstruct MRI images and estimate physical motion.
• Decompose the complex optimization problem into two computationally manageable sub-problems for efficient solving.
• Validate the CS+M framework across diverse clinical applications, including cardiac cine, cardiac perfusion, and brain perfusion imaging.

Main Results:

• Demonstrate significant reduction in blurring artifacts in reconstructed dynamic MRI images.
• Showcase the preservation of target shape and fine details, crucial for accurate medical diagnosis.
• Achieve superior reconstruction quality compared to state-of-the-art techniques, especially at high undersampling rates.

Conclusions:

• The proposed CS+M framework offers a robust and efficient solution for high-quality dynamic MRI reconstruction from undersampled data.
• The model's ability to handle motion artifacts and undersampling enhances its clinical applicability and generalizability.
• This approach represents a significant advancement in medical imaging, improving diagnostic capabilities through enhanced image fidelity.