Abstract
Purpose :
In order to derive automated methods to distinguish anisometropic amblyopia from normal patients’ eye movements, this study tests if there are significant differences in saccade latency, amplitude, and corrective saccades between anisometropic and normal eye movements. We hypothesize that anisometropic (like amblyopic) eyes will have longer saccade latency, greater saccade amplitude variability, and more corrective saccades while localizing a stimulus.
Methods :
We used the Tobii Pro Fusion eye tracker (250 Hz) to record fine-grained eye movements. The stimuli consisted of 2 cartoon images. Five participants were instructed to fixate on each of the 11 image instances (22 total instances) appearing for 2.5 seconds in a randomized location on the screen. To simulate anisometropia, participants repeated the experiment twice while wearing a single prescription plus lens with a power of +1.75 on their (1) right and (2) left eye. Saccades were extracted based on the Tobii eye tracker classification (velocity > 30 deg/s). We calculated saccade latency (time between the stimuli appearing and the primary saccade), saccade amplitude (distance between 2 fixation points in degrees), and the number of corrective saccades (number of saccades between the primary and secondary saccade).
Results :
Mann-Whitney U tests were conducted to compare normal and anisometropic eye movements. No significant differences were observed in primary and secondary saccade amplitudes or in saccade latency between normal vision, left-eye, and right-eye anisometropia. However, a significant increase in corrective saccades of 19.5% (p<0.05) emerged in the anisometropic group compared to the normal group.
Conclusions :
These findings offer partial support for our initial hypothesis: there was a significant increase in corrective saccades between anisometropic and normal eye movements. The observed significant differences in corrective saccades for anisometropic eye movements alone suggests the potential for further significant differences in truly anisometropic amblyopic eye movements. Further investigations, especially those involving a larger and more diverse sample, will enable more nuanced evaluation of these saccadic characteristics. These initial findings are valuable, as anisometropia is a risk factor for amblyopia, and they lay the groundwork for future research, including assessment of these characteristics in amblyopic vs. normally-sighted children.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.