Computer-Aided Intraoperative Toric Intraocular Lens Positioning and Alignment During Cataract Surgery

Area of Science:

• Ophthalmology
• Computer Vision
• Medical Imaging

Background:

• Cataract surgery often involves intraocular lens (IOL) implantation to restore vision.
• Toric IOLs are increasingly used to correct corneal astigmatism, requiring precise axis alignment.
• Current manual marking methods for IOL alignment can be time-consuming and less accurate.

Purpose of the Study:

• To develop and evaluate a computer-assisted framework for markerless intraoperative IOL positioning and alignment.
• To improve the accuracy and efficiency of toric IOL placement during cataract surgery.
• To provide a non-invasive alternative to conventional manual marking techniques.

Main Methods:

• Iris boundary segmentation and eye center determination using statistical sampling and deep convolutional neural networks (CNNs).
• Eye rotation tracking utilizing scleral capillaries in regions of interest (ROIs) with correlation filter and CNN methods.
• Correction of cumulative tracking errors using a reference image for enhanced long-term stability.

Main Results:

• The proposed algorithm achieved an average position error of approximately 0.2 mm.
• Axis alignment error was less than 1 degree, demonstrating high precision.
• The system operated at a frame rate exceeding 25 FPS, suitable for intraoperative use.

Conclusions:

• The developed framework enables accurate and efficient markerless IOL positioning and alignment.
• This computer-assisted approach has the potential to significantly improve outcomes in cataract surgery.
• The technology offers a faster, more accurate, and non-invasive method for toric IOL implantation.