Medical image processing method and device

Abstract

The invention discloses a medical image processing method and device, and the method comprises the steps: carrying out the Fourier transform of a medical image, and obtaining a high-frequency signal component and a low-frequency signal component in a frequency domain vector signal; performing three-dimensional filtering sharpening based on the high-frequency component to obtain a corresponding enhanced image signal, obtaining a frequency domain module value corresponding to the high-frequency component based on the high-frequency component, and calculating by using the frequency domain module value to obtain a module value map; obtaining background information data in the medical image based on the low-frequency component; and combining the enhanced image signal with the corresponding module value map, performing weighted fusion with the background information data, and calculating to obtain a target medical image.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a method and device for processing medical images. Background technique [0002] When imaging with magnetic resonance technology, the image generated by the original data will be affected by factors such as background noise, artifacts, and boundary blur effects, which will lead to distortion of the tissue structure information of the image, which will not be conducive to the subjective judgment of the doctor. The original image obtained by scanning is processed after filtering and sharpening to reduce the unfavorable image while highlighting the anatomical structure between the image tissues. [0003] Because the original MRI images often have background noise or unfavorable factors of the scanning system (such as inhomogeneity of the magnetic field, output distortion of the gradient coil due to eddy current generated by high-frequency oscillation), the tissue signal is not cle...

AI Technical Summary

Problems solved by technology

[0005] In the existing magnetic resonance image filtering algorithm, the median filter will seriously lose the high-frequency information of the signal when filtering out Gaussian noise, blurring the details such as the edge of the image; in addition, its filtering effect is often affected by the noise intensity and the filtering window. Constrained by factors such as size and shape

Cubic spline interpolation filtering (cubic interpolation) has high interpolation accuracy, smoother edges, low image loss quality, but a large amount of calculation

While traditional digital filters such as Gaussian high-pass filters focus on improving the high-frequency signal of the image, they also increase the noise signal in the image, which leads to a decrease in the signal-to-noise ratio of the image.

Two-dimensional anisotropic filtering can enhance the effect of planar image signals, but it lacks effective means to improve the contrast of tissue signals in three-dimensional space, especially in the direction perpendicular to the plane.

Images