An Efficient Lens Structures Segmentation Method on AS-OCT Images.

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Accurate segmentation of lens structures in anterior segment optical coherence tomography (AS-OCT) images is crucial for cataract grading.
• Existing methods may face challenges with computational cost and segmentation accuracy.

Purpose of the Study:

• To develop an efficient lens structure segmentation method for AS-OCT images.
• To reduce computational cost while maintaining high segmentation accuracy for cataract grading.

Main Methods:

• Utilized an efficient semantic segmentation network for lens area extraction, replacing conventional object detection.
• Applied the semantic segmentation network again for detailed lens structure segmentation.
• Incorporated curve fitting processing (CFP) to refine segmentation results.

Main Results:

• The proposed method demonstrated good performance in both accuracy and processing speed.
• Achieved efficient segmentation of lens structures in AS-OCT images.
• The method is suitable for practical applications on devices like CASIA II.

Conclusions:

• The developed efficient segmentation method offers a viable solution for AS-OCT image analysis in ophthalmology.
• The approach balances computational efficiency with high accuracy for cataract grading.
• The technique shows promise for real-world clinical applications.