X-ray Image Processing Method for Dynamic Range Expansion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current X-ray image processing methods fail to achieve smooth transitions in images, leading to loss of information in regions with low attenuation, particularly in breast examinations with large breasts, due to image saturation issues caused by limited dynamic range detectors and the absence of physical anti-scatter grids.
Innovation Solution
An X-ray image processing method that involves acquiring images with different exposure parameters, determining boundary lines and regions, normalizing pixel values, and allocating weighting factors to fuse images, thereby expanding the dynamic range and preventing saturation, as described by the formulas for weighting factors w1 and w2.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radiation dosage is increased to improve image quality in thick breast regions, then image quality in thick regions is improved, but saturation occurs in low attenuation regions causing information loss
Solution Approach 1:
The image is divided into multiple regions based on attenuation characteristics (first region with first pixel value range, second region with second pixel value range). Different processing strategies are applied to each region: the first region uses normalization based on boundary line pixel values, while the second region uses different weighting factors, thereby resolving the contradiction between improving thick region quality and preventing low attenuation region saturation
Solution Approach 2:
Different regions of the image are assigned different quality characteristics through region-specific processing. The first region (low attenuation) undergoes normalization to preserve detail information, while the second region (high attenuation) uses alternative weighting to enhance visibility, allowing each region to have optimized quality appropriate to its characteristics
2Device complexity
If detector dynamic range is small, then device complexity is reduced, but image saturation occurs in low attenuation regions
Solution Approach 1:
The patent changes the parameter ranges used for different image regions. By applying normalization with boundary-based pixel value references to the first region and different weighting factors to the second region, the effective dynamic range is expanded beyond the physical detector limitations, preventing saturation without requiring a more complex detector
3Device complexity
If physical anti-scatter grid is not used, then device complexity and radiation dose are reduced, but image quality deteriorates due to scatter radiation
Solution Approach 1:
The patent replaces the mechanical anti-scatter grid system with a computational image processing system. By using region-based normalization and differential weighting algorithms, the method achieves scatter compensation and image quality enhancement through software rather than hardware, reducing device complexity while maintaining or improving image quality
4Productivity
If image fusion is performed without region division, then processing speed is improved, but smooth transition is not achieved affecting diagnosis
Solution Approach 1:
The patent segments the image into distinct regions based on pixel value ranges before fusion. This segmentation enables the application of region-specific processing parameters that ensure smooth transitions at boundaries, maintaining diagnostic quality while allowing efficient parallel processing of different regions
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
The present invention relates to an X-ray image processing method and system and a computer storage medium. The X-ray image processing method comprises: successively acquiring first and second images of an examination subject with different exposure parameters; determining a boundary line of the examination subject in the second image, and dividing the second image into first and second regions; subjecting the first region and/or the second region to normalization on the basis of a boundary line pixel value of the examination subject; determining, in the first image, third and fourth regions corresponding to the first and second regions, according to a pixel point coordinate correspondence relationship between the first and second images; and allocating weighting factors for the first to the fourth regions, and fusing the first region with the third region and the second region with the fourth region on the basis of the weighting factors, to form a third image. The present invention can expand the dynamic range of a detector and achieve smooth transition in images, and the image finally obtained can retain complete useful information.