Abstract
The article discusses modern methods for reducing the wave heterogeneity that occurs during magnetic resonance imaging. Due to the frequent use of magnetic resonance imaging for clinical diagnostics, considerable attention is paid to the automatic analysis of the obtained images using computer vision and pattern recognition methods. When developing such computer diagnostic tools, a frequently encountered problem is the correction of intensity heterogeneity in MR images. Wave heterogeneity caused by various factors significantly complicates the automatic analysis of magnetic resonance images (MR images). The paper presents the most popular mathematical models used to correct document data, such as a low-frequency model, a hypersurface model, and a statistical model. A comparative analysis of these methods is carried out taking into account their efficiency and computational complexity. The results obtained can be used to develop more accurate and reliable methods of computer diagnostics based on MR images.