Color-Coded QTUS Breast Tissue Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer-aided detection/diagnosis systems for breast imaging face challenges in accurately classifying normal breast tissue types, which hampers the specificity of tumor detection and requires improved methods for distinguishing between different tissue types such as skin, fat, glands, ducts, and connective tissue.
Innovation Solution
The use of quantitative transmission ultrasound (QTUS) systems that generate speed of sound, attenuation, and reflection images, combined with machine learning techniques, to classify and color-code breast tissue types, enabling precise identification and visualization of tissue types through pixel-level analysis and color coding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image intensity based and shape based parameters are used for tissue classification, then the classification can be performed using established methods, but the accuracy of distinguishing between different tissue types is insufficient
Solution Approach 1:
The patent transitions from using conventional image intensity and shape parameters to using quantitative ultrasound parameters including speed of sound, attenuation, and backscatter coefficients. This parameter transformation enables more accurate differentiation between normal tissue types (fat, glandular, fibrous) and pathological tissues, directly improving classification accuracy and tumor detection specificity
Solution Approach 2:
The patent combines multiple ultrasound-derived parameters (speed of sound, attenuation, backscatter coefficients) into a composite classification approach. This multi-parameter composite method overcomes the limitations of single-parameter classification and enables more reliable distinction between different tissue types, resolving the contradiction between classification accuracy and detection specificity
2Measurement precision
If multiple ultrasound parameters are used for tissue classification, then the accuracy of tissue type identification is improved, but the complexity of the system increases
Solution Approach 1:
The patent segments the tissue classification process into distinct stages: first acquiring multiple ultrasound parameters (speed of sound, attenuation, backscatter), then processing each parameter separately through calibration and normalization, and finally integrating them for classification. This segmentation manages system complexity by breaking down the multi-parameter analysis into manageable modular components
Solution Approach 2:
The patent performs preliminary calibration and normalization of each ultrasound parameter before final classification. Speed of sound values are calibrated to known tissue references, attenuation coefficients are normalized, and backscatter coefficients are standardized. These preliminary actions prepare the data in advance, reducing the computational complexity during the actual classification phase 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
This approach enhances the accuracy of breast tissue classification, improves the detection of abnormalities, and potentially reduces unnecessary biopsies by providing detailed, color-coded visualization of tissue types, thereby improving the specificity of tumor detection.
Implementation Method 1
Speed of sound (SOS), attenuation and reflection images obtained through quantitative transmission ultrasound (QTUS) can be used to detect and determine a tissue type
Implementation Method 2
Speed of sound (SOS), attenuation and reflection images obtained through quantitative transmission ultrasound (QTUS)
Data Source
AI summary
The speed of sound, attenuation, and reflection data obtained through quantitative Transmission ultrasound (QTUS) differs by body tissue type. Skin, fat, gland, duct and connective tissues can be classified based on the sound, attenuation, and reflection data. The system can assign coloration to breast images to provide a color-coded breast tissue volume based on the output of the classifier.