Microwave Breast Image Classification Using 3D Shape and Texture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing microwave imaging techniques face challenges in accurately distinguishing between benign and malignant lesions in breast tissue due to limitations in image processing quality, necessitating improved methods for characterizing regions of interest based on shape and texture characteristics.
Innovation Solution
A method involving the extraction of shape and texture characteristics such as solidity, correlation, and busyness from microwave images, followed by classification using a naive Bayes or quadratic discriminant analysis classifier to separate benign and malignant lesions in a 3D decision hypersurface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional microwave image processing is used, then the imaging process is simple, but the accuracy in distinguishing benign and malignant lesions is insufficient
Solution Approach 1:
The patent segments the lesion characterization process into distinct feature extraction stages: shape features (solidity, extent, roundness) and texture features (entropy, homogeneity, contrast) are extracted separately and independently, then integrated for classification. This segmentation allows each feature to be processed with appropriate algorithms while maintaining overall system manageability.
Solution Approach 2:
The patent transitions from traditional 2D image analysis to 3D volumetric analysis by extracting shape features that consider three-dimensional characteristics of lesions. The use of 3D solidity, extent, and roundness metrics adds dimensional information that improves lesion characterization accuracy without proportionally increasing processing complexity.
2Reliability
If more image processing features are extracted to improve lesion differentiation, then classification accuracy improves, but processing time and computational load increase
Solution Approach 1:
The patent extracts only the most discriminative features from the full image data: six shape features (solidity, extent, roundness, compactness, sphericity, eccentricity) and six texture features (entropy, homogeneity, contrast, correlation, energy, dissimilarity). This selective extraction of critical features maintains high classification reliability while minimizing processing time compared to using all possible image features.
Solution Approach 2:
The patent transforms raw image data into standardized feature parameters with specific mathematical definitions. Each feature is calculated using standardized formulas that convert complex image patterns into comparable numerical values, enabling efficient processing while preserving diagnostic information. The features are normalized to consistent scales for uniform processing.
3Measurement precision
If detailed shape and texture analysis is performed on all regions, then lesion detection accuracy improves, but the complexity of region identification increases
Solution Approach 1:
The patent performs preliminary segmentation to identify potential lesion regions before detailed feature extraction. Regions of interest are pre-identified using thresholding and connected component analysis, which simplifies the subsequent detailed shape and texture analysis by limiting processing to only relevant areas rather than the entire image.
Solution Approach 2:
The patent applies different analysis strategies to different regions: simple thresholding for initial region identification, detailed shape feature extraction for identified lesions, and texture analysis for further characterization. Each region receives the appropriate level of analysis based on its identified characteristics, optimizing both accuracy and processing efficiency.
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
Enhances the ability to differentiate between benign and malignant lesions by leveraging a limited number of crucial shape and texture characteristics, improving the accuracy and reliability of microwave imaging for breast pathology detection.
Implementation Method 1
Microwave imaging uses emission probes configured to illuminate all or part of the organ to be imaged by means of electromagnetic waves. The emitted waves pass through the zone to be imaged and are received by receiving probes.
Implementation Method 2
The received waves have passed through the zone to be imaged having undergone reflections on the obstacles encountered, at locations of dielectric contrast (for example a cancerous lesion located in the breast tissue).
Data Source
AI summary
The invention relates to a method for processing medical images of human tissues of a zone of the body of a patient and in particular of the breast by means of a microwave medical imaging device, the method comprising the following steps implemented in a processing unit of the medical imaging device:—identifying at least one region of interest using at least one initial microwave image of a zone of the body of a patient;—processing each region of interest identified in an image so as: o to determine at least a first shape characteristic, preferably the solidity of each region of interest; o to determine at least a second and third characteristics relative to the texture of each region of interest; the first, second and third characteristics being coordinates characterising each region of interest;—locating, in a space of at least three dimensions, the dimensions of which are at least the first, second and third characteristic, respectively, each region of interest based on its coordinates, the space being partitioned by a decision hypersurface into two continuous and separate sub-spaces, one sub-space such that a region of interest located therein is associated with a benign lesion, and another sub-space such that a region of interest located therein is associated with a malignant lesion.