Deep Learning Off-Axis Scatter Suppression in Minimum Variance Ultrasound Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The performance of minimum variance beamforming methods in ultrasound imaging, particularly in terms of contrast, has not been satisfactory, and there is a need to enhance the imaging quality without compromising resolution.
Innovation Solution
A high-contrast minimum variance imaging method based on deep learning is proposed, which integrates a deep neural network to suppress off-axis scatter signals by processing echo data through delay operations, Fourier transforms, and weighted summations, improving image contrast while maintaining resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If conventional minimum variance beamforming method is used, then image resolution is maintained, but image contrast is insufficient
Solution Approach 1:
A deep neural network is introduced as an intermediary component between the raw ultrasound data and the final beamformed image. The neural network processes the channel data to generate optimized weights that suppress off-axis scatter signals, thereby improving contrast while maintaining the resolution benefits of conventional minimum variance beamforming.
Solution Approach 2:
The invention changes the weight parameters used in beamforming by employing a deep neural network to learn optimal weight values from training data. This allows dynamic adjustment of beamforming weights to maximize contrast while preserving resolution, overcoming the limitations of fixed conventional weight calculation methods.
2Illumination intensity
If deep learning is integrated to suppress off-axis scatter signals, then image contrast improves, but computational complexity increases
Solution Approach 1:
The deep neural network is trained in advance using simulated or experimental data to learn the complex patterns of off-axis scatter signals. Once trained, the network can rapidly process new ultrasound data to improve contrast without requiring complex real-time computations during actual imaging, thus reducing online computational complexity.
Solution Approach 2:
The invention uses simulated ultrasound data to train the deep neural network, creating a virtual model that replicates real-world scattering patterns. This allows the network to learn contrast enhancement techniques from synthetic data, reducing the need for extensive real-data processing and lowering overall computational complexity.
Data Source
AI summary
Disclosed is a high-contrast minimum variance imaging method based on deep learning. For the problem of the poor performance of a traditional minimum variance imaging method in terms of ultrasonic image contrast, a deep neural network is applied in order to suppress an off-axis scattering signal in channel data received by an ultrasonic transducer, and after the deep neural network is combined with a minimum variance beamforming method, an ultrasonic image with a higher contrast can be obtained while the resolution performance of the minimum variance imaging method is maintained. In the present method, compared with the traditional minimum variance imaging method, after an apodization weight is calculated, channel data is first processed by using a deep neural network, and weighted stacking of the channel data is then carried out, so that the pixel value of a target imaging point is obtained, thereby forming a complete ultrasonic image.