Automatic high-bandwidth calibration and reconstruction of arbitrarily sampled parallel MRI

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging (MRI)
• Signal Processing

Background:

• Undersampled MRI data requires advanced reconstruction techniques.
• Compressed sensing and parallel imaging (pMRI) are established methods for MRI reconstruction.
• Autocalibrating pMRI techniques simplify reconstruction by not requiring explicit coil sensitivity knowledge.

Purpose of the Study:

• To develop a novel autocalibration approach for pMRI utilizing smooth, high-bandwidth coil profiles.
• To demonstrate a parameter-free reconstruction algorithm integrating autocalibrating pMRI and compressed sensing.
• To present methods for automatic parameter estimation in MRI reconstruction.

Main Methods:

• Derivation of a new autocalibration method for pMRI to estimate and use high-bandwidth coil profiles.
• Development of a parameter-free reconstruction algorithm combining autocalibrating pMRI and compressed sensing.
• Implementation of automatic parameter estimation techniques for MRI reconstruction.

Main Results:

• High-bandwidth coil models lead to higher reconstruction accuracy compared to traditional methods.
• The automatic parameter estimation techniques yield acceptable results for MRI reconstruction.
• The proposed parameter-free algorithm successfully combines autocalibrating pMRI and compressed sensing.

Conclusions:

• The novel autocalibration approach with high-bandwidth coil models enhances MRI reconstruction accuracy.
• Parameter-free reconstruction algorithms integrating advanced techniques are feasible and effective.
• This work advances MRI reconstruction by improving accuracy and simplifying parameter selection.