Automatic intraocular lens segmentation and detection in optical coherence tomography images

Area of Science:

• Ophthalmology
• Biomedical Engineering
• Optical Engineering

Background:

• Accurate analysis of intraocular lens (IOL) geometry is crucial for understanding visual outcomes.
• Existing methods for IOL analysis can be limited in scope or accuracy.

Purpose of the Study:

• To develop and validate a novel algorithm for automatic segmentation, detection, and surface analysis of accommodative IOLs within a biomechanical eye model.

Main Methods:

• The algorithm incorporates denoising and fan correction using multi-level calibration.
• Segmentation employs an adapted Canny edge detection algorithm with automatic region of interest searching.
• Optical distortion correction via inverse raytracing and spherical fitting for geometry extraction.

Main Results:

• The algorithm successfully segmented and detected the accommodative IOL in the eye model.
• Extracted lens curvature and position with accuracy comparable to previous studies.
• Demonstrated robust performance in analyzing IOL geometry and surface characteristics in vitro.

Conclusions:

• The developed algorithm provides a powerful tool for automated IOL analysis.
• The achieved accuracy supports its potential for future clinical translation.
• Further development will focus on adapting the algorithm for clinical anterior segment Optical Coherence Tomography (OCT) devices.