Related Experiment Video
Updated: Jul 28, 2025
05:24
How to Build a Laser Speckle Contrast Imaging LSCI System to Monitor Blood Flow
Published on: November 11, 2010
23.9K
Deep-learning-based 3D blood flow reconstruction in transmissive laser speckle imaging
Optics Letters
|June 1, 2023
Summary
This study introduces a deep learning method for 3D blood flow imaging in thick tissues using laser speckle imaging (LSI). The technique enhances accuracy and resolution for deep vessels, aiding in disease diagnosis.
Area of Science:
- Biomedical Optics
- Medical Imaging
- Deep Learning Applications
Background:
- Transmissive laser speckle imaging (LSI) monitors blood flow in thick tissues.
- Multiple scattering blurs speckle images, reducing accuracy and resolution for deep vessels.
Purpose of the Study:
- To develop a deep learning strategy for high spatiotemporal resolution 3D reconstruction from single transilluminated LSI images.
- To improve structural and functional details of deep vessels without complex imaging setups.
Main Methods:
- Generated a training dataset using vessel masks and depth-dependent point spread functions (PSF) based on the correlation transfer equation.
- Employed UNet and ResNet for deblurring and depth estimation.
- Estimated blood flow using a depth-dependent contrast model.
Main Results:
- Achieved high-fidelity structural reconstruction of 3D vessels.
- Demonstrated depth-independent estimation of blood flow.
- Validated with simulated data and phantom experiments.
Conclusions:
- The proposed fast 3D blood flow imaging technique offers improved accuracy and resolution for thick tissues.
- Suitable for real-time monitoring and diagnosis of vascular diseases.
