Abstract
Purpose:
Visual search patterns of naive subjects were analyzed for a word recognition task with visual noise. Varied levels of contrast were presented by adding visual noise to word and background, simulating widespread internal visual noise found in ocular and neurological disease.
Methods:
Eleven subjects included 4 males & 7 females, ages 22-50 yr with visual acuity of 20/20. The stimulus was a 20/200 size word, 5-9 letters in length, in Arial font presented centrally in the Laser Scanning Digital Camera Stimulator (LSDC-S). A 16 word list was generated from the Low Vision Training Manual (Quillman). The visual stimuli were projected in the LSDC-S with the stimulus and retinal image focal planes linked. The background illuminance and the contrast for stimuli and background were matched to Goldmann perimetry. Pattern noise at a fixed spatial frequency was presented with the words, starting with 9% and lowered by 3% over 4 levels until the word was recognized. The trial ended if recognition did not occur in 6 sec. Time to recognition was recorded for the final trial. Fixation and scan patterns were evaluated from x,y coordinates of a retinal image feature over time. A control experiment was performed without pattern noise.
Results:
No words were recognized at noise levels > 6%. At sub-threshold levels, there was no association between word length and extent of eye movements for either the horizontal or vertical direction. At threshold, the word length was significantly associated with extent of the eye movements (R2 = 0.492, P= 0.033) for the horizontal direction. Average eye movements were 765X90 microns at threshold, while they were 668X89 microns subthreshold. Movement was more closely associated with word length than syllables. Probability of more movement at word recognition trial versus non-recognition trial: P = 0.00752. Subjects viewing words under suprathreshold conditions had average eye movements of 530X98 microns.
Conclusions:
The size of horizontal eye movements during word recognition increases as word length increases at threshold noise. There is also a significant increase in eye movement size during the noise level trial with recognition as compared to noise levels that preclude word recognition. These findings provide a signature of eye movements made during word recognition, which may improve understanding of reading with diffuse central vision loss.
Keywords: 584 low vision •
672 reading •
522 eye movements