Medical Image Subtraction Processing for Multiple Sites of Interest
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical image processing techniques struggle to generate a suitable subtraction image for multiple sites of interest in medical images, particularly in diagnosing bone metastasis, spinal canal invasion, and extraosseous masses, as they often introduce noise and fail to visualize temporal changes effectively.
Innovation Solution
An image processing apparatus and method that identifies multiple sites of interest in medical images and applies specific subtraction generation methods for each site, including noise reduction and high-density emphasis processing, to generate a suitable subtraction image for observing these sites, such as using simple subtraction for the spinal canal and high-density emphasis for bone regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single subtraction generation method is applied to the entire image, then the processing is simple and fast, but the subtraction image cannot effectively visualize multiple sites of interest with different characteristics
Solution Approach 1:
The patent divides the medical image into multiple regions of interest (ROI) based on anatomical structures such as spinal canal, vertebra, and surrounding tissues. Different subtraction generation methods are then applied to each ROI segment, allowing optimized processing for each specific anatomical region while maintaining overall image coherence.
Solution Approach 2:
The patent applies different subtraction generation methods to different regions of the image based on their specific characteristics. For example, the spinal canal region uses one method optimized for visualizing canal invasion, while the vertebral region uses another method optimized for bone metastasis detection, thereby achieving local optimization for each site of interest.
2Measurement precision
If noise reduction processing is applied to the entire subtraction image, then the overall noise is reduced, but the noise in specific regions of interest may not be effectively reduced
Solution Approach 1:
The patent segments the subtraction image into multiple ROIs corresponding to different anatomical structures. Noise reduction processing is then applied independently to each segment using appropriate algorithms for that specific region, ensuring effective noise reduction in each area without unnecessarily processing the entire image.
Solution Approach 2:
The patent applies noise reduction processing locally to each region of interest based on its specific characteristics and requirements. Different noise reduction algorithms are selected for different anatomical regions, optimizing the noise reduction effectiveness for each site while avoiding unnecessary processing in other areas.
3Measurement precision
If high density emphasis processing is applied to the entire image, then bone regions are emphasized, but soft tissue differences and other sites of interest may be obscured
Solution Approach 1:
The patent divides the image into multiple anatomical regions including bone, soft tissue, and spinal canal areas. High density emphasis processing is selectively applied only to bone regions to enhance bone metastasis visualization, while other regions are processed with different methods that preserve soft tissue contrast and other important features.
Solution Approach 2:
The patent applies high density emphasis processing locally only to bone regions where it is most needed for detecting metastasis, while other anatomical structures are processed with different algorithms that preserve their specific characteristics, thereby achieving localized optimization without obscuring other sites of interest.
4Measurement precision
If multiple subtraction generation methods are applied to different regions, then the visualization quality for multiple sites is improved, but the processing time and computational load increase
Solution Approach 1:
The patent segments the image into anatomical regions and applies appropriate subtraction generation methods to each segment in parallel. This segmentation approach allows optimized processing for each region while maintaining the ability to process multiple regions simultaneously, balancing visualization quality with processing speed.
Solution Approach 2:
The patent dynamically selects and applies different subtraction generation methods based on the specific requirements of each anatomical region. The processing system adapts its behavior according to the content being analyzed, applying more complex methods only where necessary and using simpler methods for routine areas, thereby optimizing the balance between quality and speed.
Data Source
AI summary
An image processing apparatus includes: an obtaining unit configured to obtain a first medical image and a second medical image collected from an object; an identification unit configured to identify a plurality of sites of the object included in the first medical image; and a subtraction image generation unit configured to generate a subtraction image between the first medical image and the second medical image by calculating, for each of a plurality of pixels forming the first medical image, a subtraction value with respect to a pixel, corresponding to the pixel, on the second medical image by a subtraction generation method corresponding to a site based on the result of the identification unit.