Abstract
Purpose: :
To evaluate the impact of inter visit gaze changes on reproducibility of central field thickness measurement (average thickness for the central 1 mm region of the macula) and to describe how Cirrus HD-OCT minimizes the resulting variability.
Methods: :
Twenty eyes (ten with exudative age-related macular degeneration and ten with cystoid macular edema) were scanned on two different days using Cirrus HD-OCT Spectral Domain Technology (Carl Zeiss Meditec, Dublin CA). The data was analyzed using in-house research grade software. Retinal thickness maps were constructed by automatically segmenting the ILM and RPE layers. For each eye, lateral positional offset between scans from different visits was determined by correlating vessel landmarks on vessel-enhanced OCT fundus images from two visits. Error in the central field Thickness (CFT) measurement was simulated by finding the difference between the average thickness of a 1 mm area at the center of the retinal thickness map and average thickness of a 1 mm area displaced by the estimated positional offset. Results for all eyes are displayed in a scatter plot (each disease category depicted with a different color) comparing positional offset versus CFT error.
Results: :
Preliminary results showed mean positional offset and mean absolute CFT error when not using visit-to-visit registration to be 0.37 mm and 19.1 um, respectively. The preliminary results did not show distinguishable difference in CFT error distribution between the two disease categories.
Conclusions: :
Gaze changes can lower reproducibility of thickness measurements using OCT. Cirrus HD-OCT provides a bright green spoke-shaped fixation target expanding 10 degrees in the visual field giving patients with low central vision a visible target to fixate on. However from one visit to another, patients who do not clearly see the center of the target may fixate at a different part of the target which may effect outcome measurements such as the CFT. This demonstrates the need to acquire images from the same macular location between visits independently of fixation. One strategy to accomplish this, the HD-OCT’s visit-to-visit registration capability derived from the machine’s line scanning ophthalmoscope (LSO), has been recently implemented.
Keywords: imaging/image analysis: clinical • retina • eye movements