Purchase this article with an account.
Sara Garcia, Karin Petrini, Lyndon Da Cruz, Gary S Rubin, Marko Nardini; Assessing improvements in perception afforded by retinal prostheses in multisensory tasks. Invest. Ophthalmol. Vis. Sci. 2014;55(13):5962. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess whether prosthetic vision can be combined with other sensory information to enhance performance on multisensory tasks.
Five participants (4 retinitis pigmentosa, 1 choroideremia; 49-76yrs) implanted with Second Sight’s Argus II retinal prosthesis in 2008-09, participated in this study. All had some level of bare light perception, but no measurable visual acuity, prior to surgery. Following implant, two scored reliably on the visual acuity scale with the system on (2.8 & 2.9logMAR). Participants completed three tasks: (1) A size discrimination task using vision, touch, or both. (2) A reaction time task to visual, auditory or visual-auditory stimuli (flash-beep). (3) A navigation task, using visual and/or self-motion information, to reproduce a path or complete a walked triangle.
(1) No participants showed sensory combination, as all five had significantly lower size discrimination thresholds for haptic-only than visual-only judgments. Visual-haptic (VH) performance did not significantly differ from haptic-only (H) performance [Mean correct: VH=87%; H=85.5%]. (2) Two participants of five indicated sensory combination, with faster responses to visual-auditory (VA) than unisensory (A or V) stimuli [A-VA>.028ms & V-VA>.094ms]. Two of three not showing this multisensory advantage, judged flash-beep stimuli as simultaneous when flashes preceded beeps by >.144ms (compared to .034-.036ms in other participants). (3) Three participants of four indicated sensory combination on the navigation tasks, but this worsened performance. Two were significantly further away from the correct final position with vision (V) than without (NV) during path reproduction [Mean error: V=.91m NV=.44m]. One participant’s end position was significantly further away with vision, for triangle completion [Mean error: V=1.01m NV=.49m].
The tests indicate that prosthetic vision can be used with other sensory information to enable faster perception in reaction time tasks. However, prosthetic vision did not improve accuracy in spatial tasks. This may be due to differences in spatial vs temporal resolution of the device, and to a need to learn new spatial mappings. Training participants to understand how prosthetic vision is informative about the external world may lead to perceptual improvements in the longer term. We are currently assessing the effects of training on navigation task performance.
This PDF is available to Subscribers Only