Abstract
Purpose :
People with central vision loss (CVL) adopt a preferred retinal locus (PRL), to target stimuli in tasks typically performed using the fovea in normally sighted observers. The PRL is most often measured during simple fixation tasks, but it has been reported that this PRL does not necessarily correspond to the retinal locus used in other tasks. It has been challenging, however, to measure the PRL while observers are performing visual tasks involving frequent eye movements. Here, we present a method to determine the retinal location attended to during fixations in reading. We expect this attended location to be the fovea for normal controls and ask whether the attended location in CVL subjects corresponds to the fixation PRL.
Methods :
Ten normally sighted young adults (control group), and one subject with macular degeneration have been enrolled in the ongoing study. An eye tracker system (Tobii) was used to record gaze position. Participants were asked to read sentences, taken from the MNRead test, that were presented in three lines with left and right justification. The sentences were 60 characters long and contained 10-15 words. At the beginning of each trial, the eye tracker was calibrated using a 5-point fixation task, allowing it to track the fovea in controls and the fixational PRL in patients. In each of 120 trials, a sentence was displayed while eye position data were recorded. Reading was interrupted at a random time (1200 - 2100 ms after trial start) by replacing the sentence with a blank screen for 2 sec. Participants then reported the last word that they read. We computed the “word distance” between the reported word and the word that was fixated, based on the eye position data, just before screen blanking. A word distance of zero indicates that the reported word matches the word at the fovea (controls) or fixational PRL (CVL). Positive word distances indicate reporting words after the fixated word in the sentence, and negative word distances indicate reporting words before the fixated word.
Results :
Observers in the control group reported the foveated word in more than half of the trials (52.7%). Similarly, the AMD patient reported the word at the fixational PRL location in the majority of trials (67.2%).
Conclusions :
Our preliminary results validate a method for identifying the retinal location attended to during reading and provide a method for assessing reading PRLs in patients with central vision loss.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.