POCS-enhanced correction of motion artifacts in parallel MRI

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging (MRI)
• Image Reconstruction

Background:

• MRI motion artifacts degrade image quality and diagnostic accuracy.
• Corrupted k-space data, often from multiple receiver coils, are a primary source of motion artifacts.
• Existing correction methods may not fully leverage data redundancy or prior information.

Purpose of the Study:

• To present a new method for correcting MRI motion artifacts.
• To improve the accuracy of artifact correction by utilizing data redundancy and prior object knowledge.
• To validate the method on simulated and real-world motion-corrupted MRI data.

Main Methods:

• Employed a projections onto convex sets (POCS)-based method for reconstructing sensitivity encoded MRI data (POCSENSE).
• Identified and discarded corrupted k-space samples.
• Restored artifact-free images using coil sensitivity profiles and informational redundancy from phased-array data.
• Incorporated a priori known properties of the imaged object for enhanced correction.

Main Results:

• The POCSENSE method effectively identified corrupted k-space samples.
• Artifact-free MR images were successfully restored from the remaining data.
• The inclusion of a priori information significantly improved k-space error detection and image artifact correction.
• The method demonstrated efficacy on both simulated and real motion artifacts (e.g., head motion, pulsatile flow).

Conclusions:

• The presented POCS-based method offers effective correction of MRI motion artifacts.
• The technique leverages multicoil data redundancy and prior object knowledge for improved performance.
• This approach is applicable to both full and reduced MRI datasets, enhancing its versatility.