Fast algorithm for calculation of inhomogeneity gradient in magnetic resonance imaging data

Area of Science:

• Medical Imaging
• Biomedical Engineering

Background:

• Intensity inhomogeneity is a common artifact in magnetic resonance imaging (MRI) data.
• This artifact can negatively impact the accuracy of image analysis and interpretation.

Purpose of the Study:

• To develop and implement a novel algorithm for correcting intensity inhomogeneity in MRI data.
• To evaluate the performance of the new algorithm against existing methods.

Main Methods:

• The algorithm assumes intensity inhomogeneity is multiplicative and smoothly varying.
• It calculates the inhomogeneity gradient, solves for the gradient field, and fits it to a parametric surface.
• The approach was validated on simulated and real MRI data from humans and animals.

Main Results:

• The algorithm successfully restored homogeneity in all tested MRI images.
• It demonstrated superior or comparable performance to established methods like SPM, BrainSuite, and N3 on human brain images.

Conclusions:

• The proposed algorithm offers an effective alternative for MRI intensity inhomogeneity correction.
• It is computationally efficient and applicable to both human and animal MRI data.