Accelerated dynamic MRI using patch regularization for implicit motion compensation.

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging (MRI)

Background:

• Accelerated dynamic MRI enables faster image acquisition but often suffers from undersampling artifacts.
• Motion estimation and motion compensation (ME-MC) methods are crucial for improving image quality in dynamic MRI but are computationally expensive.

Purpose of the Study:

• To introduce a computationally efficient algorithm for motion-compensated accelerated dynamic MRI.
• To address the limitations of existing ME-MC methods in terms of speed and complexity.

Main Methods:

• Developed an efficient patch smoothness regularization scheme to implicitly compensate for inter-frame motion.
• The algorithm recovers dynamic MRI data from highly undersampled measurements using a novel regularization prior.
• The proposed method alternates between an inter-patch shrinkage step and a conjugate gradient algorithm.

Main Results:

• The proposed algorithm yields reconstructions with minimal spatiotemporal blurring and motion artifacts.
• Achieved comparable or superior image quality compared to state-of-the-art ME-MC methods.
• Demonstrated significantly faster reconstruction times, approximately nine times faster than existing methods.

Conclusions:

• The presented scheme provides computationally efficient and effective motion-compensated reconstruction for dynamic MRI.
• Offers a viable and faster alternative to current computationally expensive ME-MC schemes.
• Applicable to various scenarios with significant inter-frame motion and contrast changes.