Medical Image Processing Apparatus Using Frequency Band Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical imaging technologies face challenges in achieving high-resolution images with efficient image processing, particularly in ultrasonic imaging, where image quality is degraded due to noise and harmonic components, leading to unclear edges and reduced contrast.
Innovation Solution
An image processing apparatus that segments signals into fundamental and harmonic frequency bands, estimates the point spread function (PSF), performs deconvolution, and synthesizes images using different weights based on contrast-to-noise ratio (CNR) to improve image quality and reduce noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional ultrasonic imaging is used, then imaging speed is fast and cost is low, but image quality is degraded due to noise and harmonic components
Solution Approach 1:
The patent segments the ultrasonic signal into multiple frequency bands (fundamental band and harmonic bands) using a frequency domain filter. This segmentation allows separate processing of different frequency components, enabling noise suppression while preserving diagnostic information, thus improving image quality without requiring complete system redesign
Solution Approach 2:
The patent extracts and removes harmful harmonic components from the ultrasonic signal using frequency domain filtering. By identifying and extracting specific frequency bands corresponding to noise and unwanted harmonics, the system improves image quality by eliminating degrading factors while maintaining the essential diagnostic signal
2Measurement precision
If frequency band segmentation and deconvolution are applied, then noise is suppressed and image resolution is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary frequency band segmentation and PSF estimation before the main deconvolution process. By preparing the frequency domain representation and estimating the point spread function in advance, the system optimizes the deconvolution process to achieve high resolution while minimizing overall processing time
Solution Approach 2:
The patent changes the processing domain from spatial to frequency domain, transforming the deconvolution problem into a simpler frequency domain operation. This parameter change enables more efficient processing by leveraging the properties of frequency domain filters, achieving high resolution with reduced computational complexity
3Measurement precision
If harmonic band signal is included, then contrast-to-noise ratio is improved, but image synthesis complexity increases
Solution Approach 1:
The patent dynamically adjusts the weight assigned to the harmonic band signal based on the contrast-to-noise ratio calculated from the fundamental band. This dynamic weighting allows the system to adaptively combine frequency bands, improving contrast where needed while avoiding unnecessary complexity when simple processing suffices
Solution Approach 2:
The patent applies different processing weights to different frequency bands based on local image characteristics and contrast-to-noise ratios. By assigning higher weights to harmonic components only where they improve contrast, the system achieves enhanced image quality in critical regions without uniformly increasing processing complexity across the entire image
Data Source
AI summary
An apparatus for processing a medical image includes: a receiver configured to receive a signal having a plurality of frequency bands; an image reconstructor configured to segment the signal into a first signal of a first frequency band and a second signal of a second frequency band based on a signal strength, and configured to generate a first reconstructed image of the first frequency band and a second reconstructed image of the second frequency band; and an image synthesizer configured to synthesize the first reconstructed image and the second reconstructed image.