Abstract
Purpose :
Retinitis pigmentosa (RP) is a leading cause of blindness worldwide. Retinal prostheses are microelectronic devices that perform the role of photoreceptors, thereby restoring a limited degree of vision in patients with severe RP. With recent advances in technology, visual assistive devices and applications have improved object recognition through voice guidance. The purpose of this study was to investigate whether the combination of a retinal prosthesis and a voice guidance system improved visual recognition in participants with normal vision who used a retinal prosthesis simulator.
Methods :
Ten participants with normal vision wore retinal prosthesis simulators. Images captured by a web camera attached to a head-mounted display (HMD) were processed by a computer and shown on the HMD. A commercially available application, TapTapSee®, was installed on a seventh-generation iPod Touch, which was placed on the HMD. We performed two experiments, one to determine whether a system involving voice guidance and image recognition would be useful in distinguishing between round and square plates and the second to investigate the usefulness of a voice guide in identifying an object randomly selected from among five objects that participants knew about in advance. The percentages of correct answers and response times with and without voice guidance were compared.
Results :
In experiment 1, the percentages of correct answers and response times in the two conditions did not differ significantly. On average, of the answers, 93.3% ± 8.2% were correct without voice guidance and 100% were correct with voice guidance. The mean answer times were 15.8 ± 9.8 s without voice guidance and 13.7 ± 3.2 s with voice guidance. In experiment 2, in contrast, the percentages of correct answers and answer times did differ significantly between the two conditions. On average, of the answers, 42.0% ± 25.9% were correct without voice guidance and 100% were correct with voice guidance. The mean answer times were 19.6 ± 8.0 s without voice guidance and 13.4 ± 2.8 s with voice guidance.
Conclusions :
In the present study, we revealed that the image recognition–voice guidance system could help improve visual recognition with a retinal prosthesis.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.