Automated gonioscopy photography for iridocorneal angle grading

Area of Science:

• Ophthalmology
• Medical Imaging
• Glaucoma Diagnosis

Background:

• Iridocorneal angle opening assessment is crucial for glaucoma diagnosis.
• Manual gonioscopy is the traditional method, but can be subjective.
• Automated gonioscopy offers a potential for objective and reproducible angle assessment.

Purpose of the Study:

• To evaluate the agreement between manual and automated gonioscopy for grading iridocorneal angle opening.
• To assess the diagnostic accuracy of automated gonioscopy in detecting angle closure.
• To determine inter-rater reliability for automated gonioscopy image grading.

Main Methods:

• A cross-sectional observational study involving consecutive glaucoma clinic patients.
• Manual and automated gonioscopy (NGS-1) were performed on 88 eyes.
• Iridocorneal angle opening was graded using Shaffer's classification; agreement assessed by Fleiss' kappa and AUC.

Main Results:

• Automated gonioscopy detected angle closure in 4.3% of eyes compared to 23.4% with dynamic gonioscopy.
• Agreement for angle closure diagnosis between methods was low (κ=0.09).
• Area under the curve for angle closure detection was poor (0.53), with modest inter-rater agreement (κ=0.17) for automated images.

Conclusions:

• Manual and automated gonioscopy exhibit only slight agreement in assessing iridocorneal angle opening.
• The current NGS-1 automated gonioscopy requires significant improvements for clinical application.
• Further refinement of automated gonioscopy technology and technique is necessary for reliable real-world use.