Superresolution parallel magnetic resonance imaging: application to functional and spectroscopic imaging.

Area of Science:

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

Background:

• Standard parallel MRI techniques struggle with aliasing artifacts due to intra-voxel coil sensitivity variations.
• Existing methods often limit acceleration to specific encoding dimensions, restricting flexibility.

Purpose of the Study:

• To introduce Superresolution SENSE (SURE-SENSE), a novel parallel MRI reconstruction method.
• To address aliasing artifacts by leveraging intra-voxel coil sensitivity variations for improved spatial resolution.
• To enable acceleration in all encoding directions for enhanced imaging flexibility.

Main Methods:

• Reconstruction formulated as a superresolution imaging problem, combining low-resolution images from multiple coils.
• Utilizes high-resolution coil sensitivities to reconstruct a single, higher-resolution image.
• Employs Tikhonov regularization to manage noise amplification in this ill-posed inverse problem.
• Demonstrates 2D acceleration for echo-planar functional and spectroscopic imaging in the human brain using a 32-channel coil.

Main Results:

• SURE-SENSE effectively reduces aliasing artifacts by explicitly accounting for intra-voxel coil sensitivity variations.
• Achieves acceleration through increased spatial resolution, dictated by coil sensitivity profile variations.
• Enables acceleration along all encoding directions, unlike conventional SENSE.
• Successful application in 2D accelerated echo-planar functional and spectroscopic human brain imaging.

Conclusions:

• SURE-SENSE offers a robust approach to parallel MRI reconstruction, overcoming limitations of standard techniques.
• The method is particularly advantageous for low-resolution, high-temporal-resolution modalities like functional and spectroscopic imaging.
• SURE-SENSE provides a pathway for accelerated and high-resolution neuroimaging applications.