Reconstruction by weighted correlation for MRI with time-varying gradients

Area of Science:

• Medical Imaging
• Biophysics
• Signal Processing

Background:

• Magnetic Resonance Imaging (MRI) relies on reconstruction algorithms to generate images from acquired signals.
• Existing algorithms often assume specific gradient time dependencies, limiting their applicability.
• Understanding and controlling the Point Spread Function (PSF) and static field inhomogeneities are crucial for accurate MRI.

Purpose of the Study:

• To present a general reconstruction algorithm for MRI applicable to arbitrary gradient time dependencies.
• To demonstrate the equivalence of this method to most existing MRI reconstruction algorithms.
• To provide a framework for controlling the PSF and correcting for static field inhomogeneities.

Main Methods:

• A novel reconstruction algorithm is proposed, estimating spin density via weighted correlation of the Free Induction Decay (FID) signal and phase modulation function.
• The algorithm's derivation from linearity and shift invariance conditions is mathematically proven.
• An explicit representation of the Point Spread Function (PSF) within the weighted correlation method is derived.

Main Results:

• The presented weighted correlation method is shown to be equivalent to most previously proposed MRI reconstruction algorithms.
• A method for controlling the PSF and mitigating static field inhomogeneity effects is developed using the explicit PSF representation.
• A correction method for static field inhomogeneity is proposed, with its limitations clearly defined.

Conclusions:

• The generalized reconstruction algorithm offers a unified framework for MRI reconstruction, accommodating arbitrary gradient time dependencies.
• The ability to control PSF and correct for field inhomogeneities enhances image quality and diagnostic accuracy in MRI.
• This work provides a theoretical foundation and practical methods for improving MRI reconstruction techniques.