Related Experiment Video
Updated: May 19, 2026
04:51
Quantifying Elastic Properties of Environmental Biofilms using Optical Coherence Elastography
Published on: March 1, 2024
Depth-resolved phase velocity estimation in layered tissue based on an efficient additive attention network with
Guangyu Zhang1, Jinpeng Liao1,2, Zhengshuyi Feng1
1Healthcare Engineering, School of Physics and Engineering Technology, University of York, UK.
Biomedical Optics Express
|May 18, 2026
Summary
A new deep learning method improves optical coherence elastography (OCE) by efficiently estimating tissue stiffness. This advanced phase velocity estimation network (PVNet) enhances disease diagnosis and lesion detection in vivo.
Area of Science:
- Biomedical Optics
- Medical Imaging
- Machine Learning
Background:
- Optical coherence elastography (OCE) quantifies tissue stiffness for disease diagnosis.
- Conventional OCE faces limitations in phase velocity estimation, being computationally intensive and prone to errors.
Purpose of the Study:
- To develop a unified framework for efficient, depth-resolved phase velocity estimation in OCE.
- To introduce a deep learning approach for improved accuracy and reduced computational cost.
Main Methods:
- Combined spectral analysis of complex-valued signals with a deep learning inversion network.
- Validated the framework using homogeneous and layered agar phantoms.
- Analyzed depth-dependent phase velocity gradients in vivo on human skin.
Main Results:
- The proposed phase velocity estimation network (PVNet) achieved high accuracy in agar models (MAE of 0.123 ± 0.024 m/s) and human skin (MAE of 0.145 ± 0.114 m/s).
- Demonstrated effective segmentation of depth-resolved bi-layers in OCE data.
Conclusions:
- The developed deep learning framework offers a computationally efficient and accurate method for phase velocity estimation in OCE.
- PVNet shows significant potential for clinical applications, particularly in identifying sub-surface lesions and abnormalities.
