Related Experiment Video
Updated: Apr 14, 2026
12:00
Non-invasive Parenchymal, Vascular and Metabolic High-frequency Ultrasound and Photoacoustic Rat Deep Brain Imaging
Published on: March 2, 2015
12.8K
Label-free intraoperative imaging of hemodynamics using deep learning
Yan Shi1, Denghui Zhao1, Jingyi Yu1
1School of Biomedical Engineering, Shanghai Jiao Tong University, 800 Dongchuan Road, 200240, Shanghai, China.
Biomedical Optics Express
|April 13, 2026
Summary
This study introduces a novel label-free imaging method for real-time hemodynamic assessment during surgery. The technique significantly reduces imaging time and enhances visualization of blood flow, improving surgical guidance.
Area of Science:
- Medical Imaging
- Biomedical Engineering
- Neurosurgery
Background:
- Intraoperative hemodynamic visualization is vital for surgical decisions.
- Current standard, indocyanine green fluorescence imaging (ICG-FI), has limitations including time, waiting periods, and potential reactions.
- Label-free methods like LSCI and WLI lack arterial-venous differentiation and flow directionality.
Purpose of the Study:
- To develop a label-free imaging framework for synthesizing mean transition time (MTT) maps.
- To enable real-time hemodynamic assessment, including artery-vein differentiation and flow direction inference.
- To overcome limitations of existing intraoperative imaging techniques.
Main Methods:
- A cross-modal generation framework was developed to synthesize MTT maps from laser speckle contrast imaging (LSCI) and white light imaging (WLI).
- The framework was experimentally validated in rat brains.
- The synthesized MTT maps were analyzed for vasculature delineation, artery-vein differentiation, and blood flow direction.
Main Results:
- The proposed method successfully generated clear vasculature maps.
- Accurate differentiation between arteries and veins was achieved.
- Reliable blood flow direction decoding was demonstrated.
- Total imaging time was reduced by 95.8% compared to conventional ICG protocols.
Conclusions:
- The label-free cross-modal generation framework provides a fast and efficient solution for intraoperative hemodynamic assessment.
- This approach offers contrast-free, continuous surgical guidance.
- It overcomes key limitations of ICG-FI, enhancing surgical decision-making.
