Abstract
Purpose :
Peripheral prisms (PP) are commonly used for field expansion (FE) for peripheral field loss (PFL) such as homonymous hemianopia and tunnel vision. However, the perspective view through PP is translated and rotated by the prism power which results in changes to apparent object size, location, and optic flow. Recent augmented reality (AR) smart glasses are viable alternatives to PP-FE devices; they capture wide fields of view (FoV) using an embedded camera that covers the blind fields of patients. Part of the captured blind scene can be cropped and shifted into the display on the residual seeing field as AR-FE. We developed a prism simulation tool in virtual reality (VR) to understand how AR-FE (cropped and shifted perspective) differs from PP-FE (translated and rotated perspective).
Methods :
The prism simulation tool was made in Unity game engine for Meta Quest 2 head-mounted display. It can simulate any prism power in both horizontal and oblique PP configurations. Virtual PPs were implemented with dimensions, locations, and FoVs defined by prism power to prevent diplopia and apical scotoma. While the perspective of PP-FE was translated horizontally by half the prism’s width and rotated by the prism power, part of the blind scene was cropped to the prism’s dimensions and then shifted horizontally by half of the prism’s width in AR-FE.
Results :
Perspective changes were tested in 3D environments (Fig. 1). PP-FE resulted in closer looking rotated objects in the blind side (appearing larger and rotated toward the patient than normal viewing) and may show possible volume diplopia at the apex. AR-FE shifts objects from blind side to seeing by the same amount, maintaining object size and orientation and avoiding potential diplopia.
Conclusions :
We developed a VR field expansion simulation tool and demonstrated differences of PP and AR-FE approaches. AR-FE replicates the effect of PPs prescribed for PFL by cropping and shifting the scene in the blind field without issues with viewpoint rotation and translation. While perspective differences in these two methods are noticeable, impacts on detection and collision judgment to patients with PFL need to be further investigated.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.