Related Experiment Videos
Model-based estimation of ultrasonic echoes. Part I: analysis and algorithms
1Department of Electrical and Computer Engineering, Illinois Institute of Technology, Chicago, IL 60616-3793, USA.
Summary
This study models ultrasonic echoes using Gaussian functions to improve signal estimation. Expectation-maximization algorithms offer a computationally efficient and accurate method for analyzing complex echo patterns.
Area of Science:
- Ultrasound physics
- Signal processing
- Acoustics
Background:
- Ultrasonic backscattered echoes contain critical information about reflector properties and propagation paths.
- Accurate estimation of ultrasonic echo patterns is vital for material characterization and imaging.
Purpose of the Study:
- To develop advanced algorithms for estimating parameters of superimposed Gaussian ultrasonic echoes.
- To address limitations of traditional least squares (LS) methods in handling complex echo signals.
Main Methods:
- Modeling ultrasonic echoes as superimposed Gaussian functions with unknown parameters.
- Applying maximum likelihood estimation (MLE) and least squares (LS) principles.
- Developing and implementing expectation maximization (EM)-based algorithms for echo parameter estimation.
Main Results:
- The developed EM algorithms effectively decompose complex superimposed echoes into individual Gaussian echo estimations.
- EM algorithms demonstrate superior performance over LS methods, showing better convergence and independence from initial guesses.
- The proposed method successfully resolves closely spaced and overlapping ultrasonic echoes.
Conclusions:
- Expectation maximization (EM)-based algorithms provide a robust and computationally versatile approach for ultrasonic signal estimation.
- This method enhances the accuracy and efficiency of analyzing ultrasonic echo patterns for material property determination.
- The developed algorithms offer significant improvements for applications involving complex ultrasonic signal analysis.