Image background inhomogeneity correction in MRI via intensity standardization.

Area of Science:

• Medical Imaging
• Image Processing
• Magnetic Resonance Imaging

Background:

• Background inhomogeneity is a common artifact in Magnetic Resonance Imaging (MRI).
• This artifact, also known as bias field, can significantly affect image quality and subsequent analysis.
• Existing correction methods may have limitations in simplicity or effectiveness.

Purpose of the Study:

• To present a novel, automatic, and simple strategy for correcting background intensity inhomogeneity in MRI.
• To improve the accuracy and reliability of MRI data through effective bias field correction.
• To compare the performance of the proposed method against established techniques like N3.

Main Methods:

• A standardization process transforms image intensities to a standard grayscale.
• Tissue regions are identified via thresholding on the standardized image.
• Polynomial fitting is applied to estimate and correct background intensity variations iteratively.
• Intensity scale standardization ensures unbiased correction.

Main Results:

• The proposed method effectively corrects background intensity inhomogeneity in both simulated and clinical MR images.
• Iterative refinement ensures stable correction without significant changes in tissue region extraction.
• Comparative tests suggest the method is comparable, and potentially superior, to the non-parametric non-uniform intensity normalization (N3) method.

Conclusions:

• The developed automatic strategy offers a simple yet effective solution for MRI background inhomogeneity.
• The method's performance, particularly its potential edge over N3, warrants further investigation.
• This technique has the potential to enhance the utility of MRI data in various applications.