Shear Wave Optical Coherence Elastography Imaging by Deep Learning.

Area of Science:

• Ophthalmology
• Biomedical Engineering
• Medical Imaging

Background:

• Quantifying ocular tissue mechanical properties is crucial for understanding eye diseases.
• Optical coherence elastography (OCE) uses optical coherence tomography for tissue stiffness assessment.
• Traditional OCE methods face challenges like low repeatability.

Purpose of the Study:

• To develop an optimized data processing workflow for OCE using deep learning.
• To accurately predict ocular tissue biomechanical properties.
• To enhance the clinical applicability of OCE in ophthalmology.

Main Methods:

• Developed a 3D convolutional neural network (Concentration Prediction Network - CPN) for OCE data processing.
• CPN predicts sample concentrations and calculates Young's modulus.
• Validated the method using agar phantoms and in situ porcine corneas.

Main Results:

• CPN demonstrated high accuracy in predicting agar concentrations (MAE: 0.028±0.036 training, 0.036±0.024 testing).
• Successfully mapped corneal biomechanical distribution in porcine eyes under varying intraocular pressures.
• The deep learning approach significantly improved OCE efficiency.

Conclusions:

• The integrated OCE and deep learning workflow offers a more accurate and repeatable method for assessing ocular tissue biomechanics.
• This approach holds significant potential for clinical applications in diagnosing and monitoring eye diseases.
• Deep learning enhances the efficiency and precision of optical coherence elastography for ophthalmology.