Automated Iris Segmentation from Anterior Segment OCT Images with Occludable Angles via Local Phase Tensor.

Area of Science:

• Ophthalmology
• Medical Imaging
• Glaucoma Research

Background:

• Iris morphological changes are a primary cause of angle-closure glaucoma.
• Anteriorly-bowed irises increase the risk of progression from primary angle-closure suspect (PACS) to chronic primary angle-closure glaucoma (CPCAG).
• Automated iris abnormality detection is crucial for effective glaucoma management.

Purpose of the Study:

• To present a novel method for iris region extraction.
• To apply the method to anterior segment optical coherence tomography (AS-OCT) imagery for detecting iris abnormalities in occludable iridocorneal angles.
• To evaluate the performance of the proposed method against existing techniques.

Main Methods:

• Utilized a local phase tensor-based curvilinear structure enhancement method for iris region extraction.
• Applied the method to a dataset of 200 anterior chamber angle (ACA) images obtained via AS-OCT.
• Focused on images with occludable iridocorneal angles.

Main Results:

• The proposed method demonstrated superior performance compared to existing state-of-the-art methods.
• Experimental results indicated high applicability, effectiveness, and accuracy in iris region extraction.
• Successfully identified iris abnormalities relevant to glaucoma risk.

Conclusions:

• The developed phase tensor-based method offers a significant advancement in automated iris analysis for glaucoma.
• This technique shows promise for improving the early detection and management of angle-closure glaucoma.
• Accurate iris segmentation is vital for assessing glaucoma progression risk.