Abstract
Purpose :
Peripheral prism glasses expand the field of view for patients with homonymous hemianopia (HH), but perimetry fails to evaluate effectiveness for real world mobility. We report pilot testing of a new pedestrian collision detection test, including the challenge of developing a method to keep gaze and attention toward the forward path, as would be the case in everyday walking.
Methods :
The test, presented on a large screen (104°H × 63°V), simulated walking through a busy shopping mall with multiple pedestrians moving in various directions. For each trial, there were three possibilities: pedestrian on a collision course (constant eccentricity with looming) from the left, from the right, or neither side. Iterative rounds of development and pilot testing were conducted with HH subjects. We report results from 3 phases: free gaze; following a lead child through the mall; and addition of a secondary task (ST). Collison detection rates and response times (RT) were recorded for pedestrians at eccentricities of 20° or 25°. Misses were assigned the maximum response time, 5 s (maximum time to collision).
Results :
With free gaze, HH subjects quickly learned to “beat the test” by keeping their gaze well over to the blind side in a manner impossible when walking in the real world. A lead child and walking path were added, but even with instructions to “not lose sight of the child”, blind side detection rates were still very high. A ST was then added to keep gaze and attention mostly toward the forward path, but still allow scanning. Subjects called out numbers appearing within a string of letters moving randomly (±1.25°H × ±2.5°V) behind the lead child’s head. In ST-1, each character subtended 1.5° and was presented for 1 s. Blind-side median detection (N = 6) was 50% without prisms. In ST-2, the characters were smaller (1°) and presented for a random variable duration (0.5 to 1.5 s, mean 1 s). Blind-side median detection was 0% without prisms (N = 7), suggesting ST-2 better kept gaze in the walking direction. With ST-2 and 57Δ oblique peripheral prisms, blind-side detection improved significantly to 90% (p = 0.03) and median response times improved significantly from 5.0 s without to 3.4 s with prisms (p = 0.03).
Conclusions :
Our test measures the impact of field loss and the effectiveness of field expansion aids in a task relevant to real-world walking.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.