Ultrasound Echo Texture Quantification via Local Statistical Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ultrasound imaging methods for tumor diagnosis rely heavily on subjective expert evaluation of echo texture, leading to variability in diagnostic results due to the difficulty in observing slight texture variations, necessitating a scientific and objective quantification method for improved diagnostic accuracy.
Innovation Solution
A system and method for visualizing and quantifying echo texture features using ultrasound images, involving the selection of regions of interest, calculation of local statistical parameters, and generation of echo texture indexes, which are then used to enhance diagnostic efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If expert doctors evaluate echo texture based on experience, then diagnostic process is simple, but measurement precision and reliability are low due to subjective variations
Solution Approach 1:
The patent replaces the mechanical/subjective human visual evaluation system with an automated image processing system that uses algorithms to calculate statistical parameters (mean, standard deviation, skewness, kurtosis) of echo texture. This substitution eliminates human subjectivity and significantly improves measurement precision while managing system complexity through software-based solutions.
Solution Approach 2:
The patent transforms the qualitative echo texture evaluation into quantitative analysis by changing the parameters from subjective visual assessment to objective statistical measures (mean, standard deviation, skewness, kurtosis). This parameter transformation enables precise measurement of echo texture characteristics and resolves the contradiction between simplicity and precision.
2Reliability
If subjective evaluation method is used, then device complexity is low, but diagnostic reliability and consistency vary between doctors
Solution Approach 1:
The patent replaces the unreliable human subjective evaluation system with a standardized automated image processing system. The system applies consistent algorithms to calculate statistical parameters, ensuring reliable and reproducible diagnostic results across different users and time points, thereby improving diagnostic reliability.
Solution Approach 2:
The patent segments the diagnostic process into distinct computational steps: region of interest selection, statistical parameter calculation (mean, standard deviation, skewness, kurtosis), and result interpretation. This segmentation allows each step to be standardized and automated, improving reliability while making the system complexity manageable through modular design.
3Measurement precision
If detailed statistical analysis is performed on echo texture, then measurement precision improves, but loss of time increases due to complex processing
Solution Approach 1:
The patent performs preliminary actions by pre-defining the statistical parameters to be calculated (mean, standard deviation, skewness, kurtosis) and preparing the image processing framework before actual diagnosis. This allows the system to quickly compute results when needed, reducing processing time while maintaining high measurement precision through established computational methods.
Solution Approach 2:
The patent changes the approach from comprehensive complex analysis to focused calculation of specific statistical parameters (mean, standard deviation, skewness, kurtosis). This parameter selection strategy achieves high measurement precision for echo texture characterization while minimizing processing time by avoiding unnecessary computations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention provides quantification and imaging methods and a system of the echo-texture feature, comprising: obtaining an ultrasonic image; calculating all the pixel values in a selected ROI of the ultrasonic image to obtain a regional standard deviation; excluding pixels with a pixel value smaller than a multiple of the regional standard deviation in the selected ROI; counting a set of pixels centered around a Nth pixel to gather a Nth local mean, a Nth local variance and a Nth local coefficient of variance corresponding the set of pixel values, wherein N is from 1 to the number of the pixels remaining in the selected ROI; and obtaining an echo-texture index according to the local means, the local variances, or the local coefficient of variances. The imaging of echo texture which shows the echo texture distribution of the selected ROI with a color scale changing continuously from red to blue is also included. A parameter is provided to adjust the visualization enhancement of the echo texture.