Tracking retinal motion with a scanning laser ophthalmoscope

Area of Science:

• Ophthalmology
• Biomedical Engineering
• Computer Vision

Background:

• Most people with low vision retain functional vision for daily tasks.
• Understanding eye movement patterns, specifically preferred retinal locus (PRL) dynamics, is crucial for assessing visual task efficiency in low-vision individuals.
• Traditional analysis of retinal motion from scanning laser ophthalmoscope (SLO) data is manual, time-consuming, and prone to errors.

Purpose of the Study:

• To develop and validate an automated image processing technique for analyzing retinal motion from SLO image sequences.
• To improve the speed and accuracy of analyzing eye movement patterns in low-vision subjects.
• To compare the performance of the automated technique against manual analysis methods.

Main Methods:

• Development of an automated image processing algorithm using MATLAB software.
• Experimental testing of the automated technique on subjects with normal and low vision.
• Comparison of results from the automated technique with those from traditional manual analysis.

Main Results:

• The automated image processing technique demonstrated high effectiveness and accuracy for most subjects.
• The software performed well in analyzing retinal motion patterns, including fixation, saccade, and pursuit.
• The technique's performance was suboptimal only in cases with substandard SLO image quality.

Conclusions:

• The developed automated image processing technique offers a significant improvement over manual methods for analyzing retinal motion from SLO data.
• This automated approach enhances the efficiency and reliability of studying visual task performance in individuals with low vision.
• Further refinement may be needed to address limitations related to poor image quality.