Functional Medical Image Enhancement via Difference Kernel Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Functional medical images acquired using MRI often have a lower signal-to-noise ratio (SNR) and spatial resolution compared to anatomic images, leading to noisy and misleading results when fusion methods are used to enhance them, which can obscure or degrade the functional information.
Innovation Solution
A method that involves acquiring two medical images, one weighted for functional and the other for anatomic information, generating a difference image, applying a localized kernel to improve the similarity image, and subtracting the improved difference image from the functional image to enhance the medical image while preserving the functional information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional fusion methods are used to combine functional and anatomic images, then image quality and resolution are improved, but functional information is obscured or degraded
Solution Approach 1:
The patent extracts and removes the anatomic information component from the functional image by calculating the difference between the functional image and a synthesized anatomic image. This extraction process isolates the true functional information while eliminating the degrading anatomic structures, thereby preserving functional integrity while still allowing for quality enhancement through controlled integration.
Solution Approach 2:
Instead of directly fusing anatomic and functional images (traditional approach), the patent inverts the process by subtracting the anatomic image from the functional image. This inversion allows the functional information to remain dominant while selectively incorporating anatomic details, reversing the conventional fusion hierarchy to prioritize functional preservation.
2Loss of information
If functional images are acquired using MRI, then functional information is obtained, but signal-to-noise ratio and spatial resolution are lower than ideal
Solution Approach 1:
The patent introduces a synthesized anatomic image as an intermediary element that mediates between the noisy functional image and the desired high-quality output. This intermediary serves as a template that guides the enhancement process, allowing noise reduction and resolution improvement while preserving the unique functional information through careful difference calculation and selective integration.
Data Source
AI summary
Systems and methods for generating a medical image of a subject that includes functional information. First, two medical images are acquired. One is weighted based on functional information reflecting physiological functions of the subject and the other weighted based on anatomic information of the subject. A difference image between the two images are generated. By subjecting the difference image and the second image to a localized kernel, a local similarity image is generated. Using the local similarity image, an improved difference image is generated. Lastly, by subtracting the improved difference image from the first image, an enhanced medical image that retains the functional information reflecting physiological functions of the subject is generated.