B-Mode and Doppler Fusion for Ultrasound Lesion Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ultrasound diagnostic devices face challenges in accurately distinguishing lesions or blood vessels from surrounding tissues in ultrasound images, particularly in B-mode images, due to limitations in image processing techniques.
Innovation Solution
An image processing device that combines B-mode and Doppler methods to generate ultrasound images, utilizing a trained model for machine learning to recognize lesions or blood vessels, and outputs a recognition result by processing the images with a trained model, enabling accurate distinction between these features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If B-mode ultrasound imaging is used to visualize anatomical structures, then image quality and anatomical detail are improved, but the ability to distinguish blood flow and lesions is insufficient
Solution Approach 1:
The patent combines B-mode ultrasound images with Doppler blood flow information into a unified image processing input. The B-mode provides anatomical structure while Doppler adds blood flow velocity and direction data, merging these complementary information sources to enable accurate lesion recognition that neither modality could achieve alone.
Solution Approach 2:
The trained machine learning model acts as an intermediary that processes both B-mode and Doppler data, transforming them into a unified representation that highlights lesion characteristics. This intermediary processing layer integrates multiple information types and outputs enhanced recognition results that distinguish lesions from surrounding tissues.
2Loss of information
If Doppler method is used to detect blood flow, then blood flow information is obtained, but anatomical structure visualization is compromised
Solution Approach 1:
The patent merges B-mode anatomical imaging with Doppler blood flow detection into a single integrated processing framework. By combining these two modalities, the system recovers both anatomical structure detail and blood flow information simultaneously, overcoming the limitation where using one method compromises the other.
Solution Approach 2:
The patent adds the dimension of blood flow velocity and direction information to the traditional B-mode anatomical image. This multi-dimensional approach allows the system to visualize both structural anatomy and functional blood flow characteristics in the same image space, enabling comprehensive lesion assessment.
3Measurement precision
If machine learning processing is applied to ultrasound images, then recognition accuracy is improved, but processing time and computational complexity increase
Solution Approach 1:
The patent performs preliminary processing by combining and pre-processing B-mode and Doppler data before feeding them to the machine learning model. This preliminary integration of multiple data sources prepares the input in an optimized format, reducing the computational burden during actual recognition and enabling faster processing while maintaining high accuracy.
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
The solution provides highly accurate recognition of lesions or blood vessels in ultrasound images, enhancing diagnostic accuracy by distinguishing them from other tissues, and supports real-time display and storage of recognition results.
Implementation Method 1
an ultrasound probe that emits ultrasound and detects a reflected wave of the ultrasound
Implementation Method 2
blood flow distribution information obtained by a Doppler method
Data Source
AI summary
The image processing device includes a processor. The processor acquires an ultrasound image which is obtained by a B-mode method in an ultrasound diagnosis for a subject and in which an observation target region of the subject is shown, and blood flow distribution information obtained by a Doppler method used in combination with the B-mode method in the ultrasound diagnosis. The processor outputs a recognition result obtained by executing recognition processing of inputting the ultrasound image and the blood flow distribution information to a trained model to cause the trained model to recognize the observation target region.