A novel 3D Cartesian random sampling strategy for Compressive Sensing Magnetic Resonance Imaging.

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging
• Signal Processing

Background:

• Accelerated Magnetic Resonance Imaging (MRI) is crucial for reducing scan times.
• Compressive Sensing (CS) enables faster MRI by reconstructing images from undersampled data.
• Standard CS-MRI acquisition strategies can lead to aliasing artifacts and image smoothing.

Purpose of the Study:

• To propose novel acquisition strategies for accelerated 3D Compressive Sensing MRI (CS-MRI).
• To reduce aliasing artifacts and improve image quality in accelerated 3D CS-MRI.
• To enhance the achievable acceleration factors in 3D CS-MRI.

Main Methods:

• Development of two 3D Cartesian sampling strategies with random switching of the frequency encoding direction.
• Strategy 1: Swapping frequency encoding with one phase encoding direction.
• Strategy 2: Randomly selecting frequency encoding from all K-space directions.

Main Results:

• Simulations demonstrated superior performance compared to standard CS acquisition.
• Reduced aliasing artifacts and improved image quality after CS reconstruction.
• Demonstrated potential for greater achievable acceleration factors.

Conclusions:

• The proposed random switching acquisition strategies enhance 3D CS-MRI.
• These methods offer improved image quality and higher acceleration potential.
• Outperforms standard CS acquisition in simulations for accelerated 3D MRI.