Related Experiment Video
Updated: Oct 23, 2025
09:31
High-Speed Ultraviolet Photoacoustic Microscopy for Histological Imaging with Virtual-Staining assisted by Deep Learning
Published on: April 28, 2022
3.2K
Subwavelength ultrasonic imaging using a deep convolutional neural network trained on structural noise.
Yongxing Cai1, Yongfeng Song1, Peijun Ni2
1School of Traffic and Transportation Engineering, Central South University, Changsha, Hunan 410075, China.
Ultrasonics
|August 19, 2021
Summary
Subwavelength ultrasonic imaging (SUI) can now better detect flaws using a new convolutional neural network (CNN). This AI approach effectively distinguishes flaw signals from noise, improving defect detection accuracy.
Area of Science:
- Non-destructive testing
- Ultrasonic imaging
- Artificial intelligence in engineering
Background:
- Subwavelength ultrasonic imaging (SUI) enables detection of flaws smaller than the diffraction limit.
- Low signal-to-noise ratio (SNR) in SUI C-scans can hinder clear flaw visualization.
- Distinguishing flaw echoes from structural noise remains a challenge in SUI.
Purpose of the Study:
- To develop a convolutional neural network (CNN) for subwavelength ultrasonic imaging (SUI) that accounts for structural noise.
- To enhance the ability of SUI to distinguish flaw echoes from structural noise, especially at low SNR.
- To improve the accuracy and reliability of detecting subwavelength flaws.
Main Methods:
- A novel convolutional neural network (CNN) architecture was designed for SUI.
- The CNN incorporates a regression component to learn structural noise features.
- A learnable soft thresholding layer was implemented for classification of flaw echoes.
- The method was tested on artificial and natural flaws in various materials.
Main Results:
- The proposed CNN method demonstrated high performance in imaging subwavelength flaws of varying depths and sizes.
- An F1 score of 97.69 ± 1.56% was achieved in flaw detection, outperforming a time-dependent threshold method.
- Successful SUI was performed on natural flaws in spheroidal graphite cast iron without a theoretical backscattering model.
- The network showed robustness against noise distribution, multiple scattering, and complex microstructures.
Conclusions:
- The developed CNN-based SUI method effectively distinguishes flaw echoes from structural noise, significantly improving imaging quality.
- This approach offers a robust solution for subwavelength flaw detection, applicable to complex materials and conditions without prior flaw echo data.
- The method shows broad applicability for non-destructive evaluation where traditional ultrasonic imaging faces limitations.
Related Concept Videos
01:17Ultrasonography
6.8K
Ultrasonography is an imaging technique that uses high-frequency sound waves to visualize the body's internal structures. It is a non-invasive and safe procedure that does not involve the use of ionizing radiation, making it widely used in various medical fields. Ultrasonography is used to study heart function, blood flow in the neck or extremities, certain conditions such as gallbladder disease, and fetal growth and development.
During an ultrasonography procedure, a handheld device called...
During an ultrasonography procedure, a handheld device called...
6.8K
01:24Imaging Studies II: Ultrasonography
89
IntroductionUltrasonography, or renal ultrasound, is a noninvasive medical imaging technique that uses high-frequency sound waves to visualize the kidneys, ureters, bladder, and surrounding tissues.Indications for Urinary System UltrasonographyUrinary system ultrasonography is indicated in various clinical scenarios, such as:Kidney Stones (Urolithiasis): To detect and monitor the size and presence of kidney or urinary tract stones.Hydronephrosis: To assess the dilation of the renal pelvis and...
89