Medical Image Processing Apparatus for Coronary Artery Disease Diagnosis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging techniques, particularly in coronary artery disease diagnosis using CT scans, face challenges in accurately discriminating between contrast-enhanced blood vessels and calcium deposition or stent areas due to overlapping Hounsfield values, leading to difficulties in visualizing and segmenting blood vessels, especially when calcium depositions are heavy or when blooming artifacts are present.
Innovation Solution
An image processing apparatus that generates a joint histogram from aligned non-contrast and contrast images, classifies image areas into lumen, vessel wall, blooming, and misalignment regions, and uses a color model to visually differentiate these areas, enhancing the accuracy of blood vessel segmentation and reducing the impact of artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If a contrast CT scan is used to enhance blood vessel visibility, then the pixel value of blood vessel area increases, but the Hounsfield value overlaps with calcium deposition area and stent area making visual discrimination difficult
Solution Approach 1:
The patent segments the contrast image into multiple image areas (blood vessel area, calcium deposition area, stent area, etc.) based on pixel value characteristics. By dividing the image into distinct segments with different processing, each area can be optimized for its specific properties, allowing blood vessels to be enhanced while calcium and stents are suppressed without their Hounsfield values overlapping.
Solution Approach 2:
The patent applies different processing characteristics to different regions of the image. Blood vessel areas receive contrast enhancement to increase visibility, while calcium deposition areas and stent areas receive suppression processing to reduce their prominence. This local differentiation resolves the contradiction by making each region have the quality needed for its specific diagnostic purpose.
2Illumination intensity
If standard window level or color mapping technique is used to display contrast image, then the image can be visualized, but it is difficult to visually discriminate contrast-enhanced blood vessel area from calcium deposition area or stent area
Solution Approach 1:
The patent assigns different color characteristics to different image areas based on their classification. Blood vessel areas, calcium deposition areas, and stent areas are displayed with distinct colors or color intensities. This color differentiation allows all areas to be visualized simultaneously while maintaining clear visual discrimination between them, resolving the contradiction between visualization and discrimination.
3Object-generated harmful factors
If subtraction processing is used to remove common features, then blood vessel area enhanced with contrast medium can be left, but accurate alignment and classification of multiple image areas is required
Solution Approach 1:
The patent performs preliminary classification of image areas into blood vessel areas, calcium deposition areas, stent areas, and other tissues before executing subtraction processing. By pre-segmenting the image based on pixel value characteristics and anatomical knowledge, the subsequent subtraction operation becomes simpler and more accurate, as it only needs to operate on pre-identified regions rather than the entire image.
Solution Approach 2:
The patent introduces an intermediate classification step that acts as a mediator between the raw contrast image and the final subtraction result. This classification layer identifies and tags different tissue types, allowing the subtraction processing to selectively remove bone and soft tissue while preserving enhanced blood vessels, without requiring complex real-time alignment adjustments.
4Measurement precision
If manual interpretation is used to diagnose coronary artery disease, then diagnostic accuracy can be maintained, but time consumption and reliance on expert interpretation increase
Solution Approach 1:
The patent implements automated classification and processing algorithms that perform diagnostic functions previously requiring manual expert interpretation. The system automatically identifies blood vessel areas, enhances contrast, suppresses calcium and stent artifacts, and generates diagnostic-ready images. This self-service capability maintains diagnostic accuracy while eliminating the time consumption associated with manual interpretation.
Solution Approach 2:
The patent transforms the diagnostic process by changing key parameters from manual to automated operations. Image processing parameters such as contrast enhancement levels, suppression thresholds for calcium and stents, and classification criteria are optimized and applied automatically. This parameter transformation enables the system to replicate expert diagnostic accuracy through automated image processing and analysis.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
According to one embodiment, a medical image processing apparatus includes a storage unit (40), a generation unit (50), and a segmentation unit (52). The storage unit (40) configured to store a first medical image and second medical image regarding the same imaging region imaged at different times. The generation unit (50) configured to generate numerical data representing a frequency of a combination of a pixel value of the first medical image and a pixel value of the second medical image at the same coordinates. The segmentation unit (52) configured to divide the second medical image into a plurality of image areas by using the generated numerical data.