Purchase this article with an account.
Thomas Kuyk, Lei Liu, Elizabeth Belleau, Peter Smith, Leon N McLin; A Feature Search Method for Mapping the Central Visual Field. Invest. Ophthalmol. Vis. Sci. 2014;55(13):4118.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To determine if small areas of vision loss (scotoma) of approx.1-2° in the central 10-20° of the visual field could be detected using a visual feature search (FS) paradigm. The rationale was that if a scotoma (S) covered the search target or partially obscured it, response time for target detection (RTHit) would be increased compared to targets presented outside the S; either because eye movements would be needed or reduced conspicuity would lead to longer decision times.
Nine subjects completed a FS task with and without a simulated S present in central 10° or 20° search areas. Parafoveal S were simulated with a gaze-contingent display system. The FS task consisted of a black target presented in a set of grey distracters. Stimulus items were randomly distributed at the intersections of virtual square grids laid over the search areas. Distractor density was set at 20% of possible grid locations and over trials the target appeared at each test location. On 25% of trials no target was presented. S were circular with Gaussian blurred edges and centered in open areas adjacent to grid intersections. For the 10 and 20° search areas, one grid size (11 x 11; 13 x 13), three S eccentricities (2, 3 & 6°; 5, 7 & 9°) and two S sizes (1.1 & 1.6°; 1.6 & 2.2°) were tested for each search area. Each trial was subject initiated after central fixation was achieved. Subjects responded target present or target absent with button presses that also terminated trials.
For the 10° search area, a 1.1° S could be detected on the basis of increased RTHit for targets presented immediately adjacent to the S compared to RTHit averaged for all grid locations. The larger 1.6° S yielded longer RTHit and larger differences compared to average RTHit. For the 20° search area, a 2.2° S was detectable at all locations using the reference versus RTHit average. Fig 1 shows RTHit (ms) for each grid location for a 10 x10° search area. A S was located adjacent to the intersection of row 2 and column 2 and RTHit are longest in this vicinity.
The results suggest the FS paradigm has sufficient merit as a method to detect small central S to warrant further study. Future research should focus on ways to improve the methodology, reduce noise and speed data collection.
This PDF is available to Subscribers Only