Detecting glaucomatous change in visual fields: Analysis with an optimization framework

Area of Science:

• Ophthalmology
• Medical Imaging
• Data Science

Background:

• Glaucomatous progression detection is crucial for effective glaucoma management.
• Standard Automated Perimetry (SAP) is the gold standard for assessing longitudinal visual fields.
• Existing methods for progression detection can be complex and computationally intensive.

Purpose of the Study:

• To develop an efficient optimization framework for detecting glaucomatous progression from longitudinal visual field data.
• To create a method that is simpler, faster, and more efficient than previous approaches.
• To provide a tool for patient monitoring in clinics and participant classification in research.

Main Methods:

• Formulated progression detection as a convex optimization problem using longitudinal visual field data from glaucoma patients.
• Derived a representative progression direction vector from the population data.
• Applied post-hoc analysis of visual fields along the derived vector for optimal change detection.

Main Results:

• The proposed optimization method demonstrated slightly higher sensitivities at the highest specificities compared to other progression detection methods and linear regression.
• The approach is simpler, faster, and more efficient than previously developed machine learning-based methods.
• The method provides a valuable tool for identifying glaucomatous changes.

Conclusions:

• The developed convex optimization framework offers an efficient and effective method for detecting glaucomatous progression.
• This approach simplifies and accelerates the analysis of longitudinal visual field data.
• The method has significant potential for clinical patient monitoring and research applications in glaucoma.