Classification of primary angle closure spectrum with hierarchical cluster analysis

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Primary angle closure is a significant cause of irreversible blindness worldwide.
• Anterior segment optical coherence tomography (ASOCT) provides detailed imaging of ocular structures.
• Understanding anatomical variations in primary angle closure is crucial for diagnosis and management.

Purpose of the Study:

• To classify primary angle closure subtypes using ASOCT imaging.
• To correlate these classifications with disease subtypes: primary angle closure suspect (PACS), primary angle closure glaucoma (PACG), acute primary angle closure (APAC), and fellow eyes of APAC.
• To identify predominant anatomical factors within each subtype.

Main Methods:

• A cross-sectional study involving 248 eyes from 198 subjects.
• Utilized ASOCT imaging, gonioscopy, and A-scan biometry for comprehensive eye examinations.
• Employed agglomerative hierarchical clustering based on ASOCT parameters to categorize eyes.

Main Results:

• Three distinct clusters were identified based on ASOCT parameters, showing significant differences in anterior segment features.
• Cluster 1: Smallest anterior chamber depth (ACD) and area, greatest lens vault.
• Cluster 2: Thickest iris, largest iris area, deepest ACD.
• Cluster 3: Intermediate features with higher iris curvature.
• Significant differences in cluster distribution were observed among PACS, PACG, APAC, and fellow eyes, with specific clusters predominating in certain subtypes.

Conclusions:

• Hierarchical cluster analysis successfully identified three distinct clusters based on ASOCT features.
• The predominant anatomical components vary significantly among different subtypes of primary angle closure.
• These findings highlight the utility of ASOCT in characterizing and differentiating primary angle closure subtypes.