Binocular rivalry with peripheral prisms used for hemianopia rehabilitation

To determine the relative binocular signal strength of moving images that are peripherally viewed through a monocular field expansion prism as opposed to moving images viewed directly. We hypothesised that prism blur might make prism images predominate less than images viewed directly with the other eye.

We employed the binocular rivalry paradigm to measure the relative binocular effectiveness of directly viewed vs prism images. Four normally-sighted subjects tracked the rivalrous visibility of opponent-coloured targets seen dichoptically in the same part of the retinal visual field, using monocular field expansion prisms to produce the dichoptic display. We analysed the effects of external signal strength (whether or not motion was present in either image), retinal position or eccentricity of the targets, and controlled for target saturation.

We found that prism images predominate less than directly viewed images. When both eyes were presented with pattern in the dichoptic display, direct-to-prism predominance was 51%:31%. When only the direct view was presented with pattern, direct-to-prism predominance was 74%:12%; when only the prism view was presented with pattern, direct-to-prism predominance was 25%:58%. Dominance durations followed established binocular rivalry rules.

The prism image in a monocular, peripheral field expansion prism is perceptually weaker than the corresponding direct image in the other eye. However, the prism image is still seen a significant proportion of the time, especially when no moving pattern is present in the direct view. We conclude that the rivalry ratio of the prism device is sufficiently effective for clinical applications.