Ultrasound Image Noise Removal via Binarization and Compounding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Ultrasound scanning systems face challenges in producing accurate compounded images due to noise, such as side lobes and artifacts, from ultrasound signals emitted from different angles, which affect the image quality.
Innovation Solution
The method involves emitting and receiving ultrasound signals from multiple angles, converting them into input images, binarizing these images to detect noise, removing noise from affected images, and then compounding the cleaned images into a single output image using a system with conversion, binarization, comparison, noise filtering, and image compounding units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ultrasound signals are emitted from different angles to obtain compounded images, then image resolution is improved, but noise (side lobes, artifacts) is introduced affecting image accuracy
Solution Approach 1:
The patent applies preliminary action by binarizing ultrasound images before compounding to detect and remove noise. The method converts ultrasound images to binarized form, compares them to identify noise regions, removes the noise, and then performs compounding. This preliminary noise removal step ensures that the compounding process works with clean data, resolving the contradiction between achieving high resolution through multi-angle compounding and avoiding noise introduction.
2Device complexity
If noise filtering is performed after compounding, then processing is simpler, but noise affects the compounded image accuracy
Solution Approach 1:
The patent demonstrates that performing noise filtering before compounding, though appearing more complex, actually simplifies the overall processing by preventing noise propagation. The binarization and noise removal steps identify and eliminate noise in individual images before they are compounded, ensuring that the compounding operation itself doesn't need to handle noisy data, thereby maintaining image accuracy.
3Measurement precision
If multiple ultrasound images are compounded to improve resolution, then image quality is enhanced, but noise from individual images propagates to the compounded image
Solution Approach 1:
The patent applies preliminary action by implementing a noise detection and removal mechanism before the compounding process. The method binarizes multiple ultrasound images, compares them to identify consistent noise patterns, removes the detected noise from each image, and then performs compounding. This ensures that only clean, reliable data is compounded, enhancing both resolution and image quality reliability.
Solution Approach 2:
The patent employs feedback by using comparison of binarized images to detect noise. The system compares multiple binarized ultrasound images, identifies regions that differ abnormally (indicating noise), and uses this feedback information to guide the noise removal process. This feedback mechanism ensures that noise is accurately identified and removed before compounding, maintaining high image quality reliability.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively improves the accuracy of the ultrasound output image by removing noise before compounding, thereby enhancing the overall image quality.
Implementation Method 1
emitting N sets of ultrasound signals onto a target from N different angles and receiving the N sets of ultrasound signals reflected and/or scattered by the target
Data Source
AI summary
An ultrasound scanning method includes the steps of emitting N sets of ultrasound signals onto a target from N different angles and receiving the N sets of ultrasound signals reflected and/or scattered by the target; converting the N sets of ultrasound signals into N ultrasound input images; performing a binarization algorithm for the N ultrasound input images to obtain N binarized images; performing a comparison process on the N binarized images to determine whether a noise exists in at least one of the N binarized images; when the noise exists in an i-th binarized image of the N binarized images, removing the noise from an i-th ultrasound input image corresponding to the i-th binarized image; and compounding the N ultrasound input images into an ultrasound output image.