LAUSR

Quantifying ocular tissue mechanical properties is pivotal for elucidating eye disease etiology and progression. Optical coherence elastography (OCE), leveraging high‐resolution optical coherence tomography, promises tissue stiffness assessment. Traditional OCE relies on data processing of the time‐of‐flight method and encounters challenges like low repeatability. Our study presents an optimized data processing workflow integrating OCE with deep learning to predict ocular tissue biomechanical properties. The concentration prediction network (CPN), a 3D convolutional neural network, predicts sample's concentrations and calculates the Young's modulus based on the relationship between agar concentration and Young's modulus from mechanical testing. The CPN showed high accuracy, with a mean absolute error of 0.028 ± 0.036 for training and 0.036 ± 0.024 for testing data of agar phantoms. In situ porcine corneas with various intraocular pressures was measured, yielding corneal biomechanical distribution via deep learning method. This approach enhances the efficiency of OCE and underscores potential clinical applications in ophthalmology.