A unified framework for glaucoma progression detection using Heidelberg Retina Tomograph images

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Science

Background:

• Glaucoma is a leading cause of blindness globally, characterized by optic nerve head (ONH) changes.
• Monitoring glaucomatous progression is crucial for managing primary open angle glaucoma (OAG).
• Ocular imaging, like the Heidelberg Retina Tomograph (HRT), provides quantitative ONH topography measurements.

Purpose of the Study:

• To present a novel framework for detecting glaucomatous progression using HRT images.
• To improve the accuracy of ONH change detection by incorporating spatial pixel dependencies.
• To apply the Variational Expectation Maximization (VEM) algorithm for inferring topographic ONH changes.

Main Methods:

• Development of a new framework for glaucoma progression detection using HRT images.
• Application of Markov Random Fields to model spatial pixel dependency in change detection maps.
• Utilizing the Variational Expectation Maximization (VEM) algorithm for topographic ONH change inference.

Main Results:

• The proposed framework integrates a priori knowledge, specifically spatial pixel dependency, into the change detection map.
• This is the first known application of the VEM algorithm within a glaucoma progression detection framework.
• The diagnostic performance of the new framework is evaluated against existing progression detection methods.

Conclusions:

• The novel framework enhances glaucoma progression detection by leveraging spatial information in HRT images.
• The integration of Markov Random Fields and VEM offers a more robust approach to monitoring ONH changes.
• This work contributes to more accurate and timely management of glaucoma patients.