Functional Performance of a Vibro-tactile Sensory Substitution Device in People with Profound Vision Loss

Abstract

Purpose : Reduced social engagement is a significant issue for people with vision loss. Traditional aids such as long canes and guide dogs are still the mainstay of support for mobility but are unable to detect people and faces. Emerging technologies such as sensory substitution devices (SSDs) are gaining wide-spread interest in the low vision community and may be able to provide cues for detection of people. The aim of this project was to assess the efficacy of a prototype vibro-tactile SSD to enable face detection for people with profound vision loss.

Methods : The face detection task was a forced choice design, where the participants were asked to step toward the first face they detected using the device, and then count the number of faces present. Retro-reflective markers were placed on participants’ bodies to track their movement using a Vicon system (Oxford, UK), to give a measure of reaction time and location accuracy. Qualitative exit interviews provided user experience data.

Results : Data were collected from 7 people with low vision and 4 orientation and mobility experts with simulated low vision. All participants had natural or occluded visual acuity of light perception or worse. Participants showed a significant improvement in face detection rate (p=0.025) when the device was on, compared to the device was off (fig 1). As the task only required the participants to step towards faces that they could detect, the movement analysis is only for trials when the device was on. The average response time was 10.5 s ± 5.9 s. The average first stride length was 43.8 cm ± 12.6 cm. The average angular difference between participants’ first response step and the identified face direction was within 20 degrees (18.9 degrees ± 5.9 degrees).

Conclusions : This study has shown the vibro-tactile SSD prototype has the potential to improve functional outcomes (i.e., face detection), even with brief training (<10 minutes). Qualitative interviews with the participants found that they felt more training would aid them to better interpret the SSD spatial information and improve localization. Future studies will look at training requirements for this tactile SSD.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

 

Figure 1: Percentage of trials where the number of faces present was correctly identified. 20% Device OFF trials were zero faces, so the 20% success rate with device OFF is misleading. LV= low vison participant, OM= orientation and mobility instructor