Tumor Feature Quantification via Moving Variance Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ultrasonic imaging technologies for tumor diagnosis rely heavily on subjective human input for tumor contour identification, leading to inconsistent and unreliable diagnoses due to the reliance on manual input and the blurriness of tumor margins, which affects the accuracy of feature identification such as calcifications, cysts, and heterogeneities.
Innovation Solution
A computer-based method that processes gray-scale ultrasonic images to objectively quantify and image tumor features like margins, cysts, calcifications, and heterogeneities by retrieving tumor contours, calculating gradient values, and defining threshold ranges to identify and display these features, providing a more reliable diagnostic tool.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual tumor contour identification is used by doctors, then the diagnosis process is simple and quick, but the reliability and consistency of diagnosis deteriorate due to subjective variability
Solution Approach 1:
The system enables automatic self-identification of tumor contours through image processing algorithms. The computer automatically retrieves tumor contours from ultrasonic images without requiring manual intervention, thereby eliminating subjective variability while maintaining operational efficiency.
Solution Approach 2:
The manual mechanical process of contour drawing by doctors is replaced with an automated image processing system. The computer uses algorithms to automatically identify and retrieve tumor contours from ultrasonic images, substituting human subjective judgment with objective computational analysis.
2Reliability
If snake algorithms are used for automatic contour identification, then objectivity is improved, but the accuracy deteriorates when tumor margins are blurred
Solution Approach 1:
The system changes the parameters used for contour identification by incorporating multiple image processing techniques beyond simple gradient-based methods. It uses region-based segmentation, texture analysis, and multiple thresholding strategies to accurately identify contours even when margins are blurred, thereby improving measurement precision while maintaining objectivity.
3Reliability
If quantitative analysis of tumor features is implemented, then diagnostic reliability is improved, but the complexity of the diagnosis process increases
Solution Approach 1:
The system achieves multi-functionality by integrating multiple diagnostic capabilities into a single platform. It simultaneously performs contour retrieval, cyst identification, calcification detection, and heterogeneity analysis, thereby improving diagnostic reliability without proportionally increasing system complexity through unified processing architecture.
Solution Approach 2:
The complex diagnostic process is segmented into distinct modular components: contour retrieval module, cyst detection module, calcification detection module, and heterogeneity analysis module. Each module handles a specific task independently, making the overall complex system manageable and maintainable while providing comprehensive quantitative analysis.
Data Source
Figure 1A~1B
Figure 2
Figure 3A
AI summary
A quantification method and an imaging method are disclosed, capable of quantifying the margin feature, the cysts feature, the calcifications feature, the echoic feature and the heterogenesis feature of a tumor, and capable of imaging the margin feature, the cysts feature, the calcifications feature and the heterogenesis feature of a tumor. The quantification method and the imaging method calculate the moving variance of the gray scale of each of the pixel points based on the gradient value of the gray scale of these pixel points. Then, depending on the purpose of the quantification method or the imaging method, the maximum value, the minimum value, the mean value, and the standard deviation of the moving variance of the gray scale of these pixel points are calculated, respectively. At final, with the definition of the threshold value and the imaging rule, the above features of the tumor are quantified or imaged.