June 2023
Volume 64, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2023
Pedestrian collision avoidance test for peripheral field loss using head-mounted display
Author Affiliations & Notes
  • Jae-Hyun Jung
    Ophthalmology, Harvard Medical School, Boston, Massachusetts, United States
    Schepens Eye Research Institute of Massachusetts Eye and Ear, Boston, Massachusetts, United States
  • Sujin Kim
    Ophthalmology, Harvard Medical School, Boston, Massachusetts, United States
    Schepens Eye Research Institute of Massachusetts Eye and Ear, Boston, Massachusetts, United States
  • Jonathan Doyon
    Ophthalmology, Harvard Medical School, Boston, Massachusetts, United States
    Schepens Eye Research Institute of Massachusetts Eye and Ear, Boston, Massachusetts, United States
  • Sandhya Shekar
    Ophthalmology, Harvard Medical School, Boston, Massachusetts, United States
    Schepens Eye Research Institute of Massachusetts Eye and Ear, Boston, Massachusetts, United States
  • Alex Hwang
    Ophthalmology, Harvard Medical School, Boston, Massachusetts, United States
    Schepens Eye Research Institute of Massachusetts Eye and Ear, Boston, Massachusetts, United States
  • Footnotes
    Commercial Relationships   Jae-Hyun Jung None; Sujin Kim None; Jonathan Doyon None; Sandhya Shekar None; Alex Hwang None
  • Footnotes
    Support  NIH R01 EY31777
Investigative Ophthalmology & Visual Science June 2023, Vol.64, 1980. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jae-Hyun Jung, Sujin Kim, Jonathan Doyon, Sandhya Shekar, Alex Hwang; Pedestrian collision avoidance test for peripheral field loss using head-mounted display. Invest. Ophthalmol. Vis. Sci. 2023;64(8):1980.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : Safe mobility is important for peripheral field loss. However, evaluating collision detection and avoidance with patients in a real-world obstacle course is challenging due to safety concerns. Previously, a video-based collision detection test was developed, showing first-person perspective walking videos to a standing subject on a large screen with a guided fixation. However, the setup lacked critical clues (e.g., gaze movement, depth, and locomotive sensory cues) utilized in natural collision avoidance behaviors. We developed a new virtual reality (VR) collision avoidance test on the Meta Quest 2 head-mounted display (HMD), which enabled these limited cues for natural collision detection and avoidance with actual walking and gaze scanning.

Methods : Subjects with homonymous hemianopia (HH; n=6) and normal vision (NV; n=8) walked in an empty real-world path while various VR pedestrian collision events occurred in a busy shopping mall on the HMD. Subjects were asked to detect and avoid 40 collision events. Colliding pedestrians approached from ±20°, ±40° within HMD’s field of view (FoV), or ±60° (out of HMD FoV) bearing angles. Ten non-colliding pedestrians were also presented in each trial. Subjects pressed a button on the colliding pedestrian’s approaching side as soon as they detected a potential collision, then avoided them naturally (e.g., changing path or speed). We analyzed detection rate, response time, number of collisions, and head rotations.

Results : HH showed a 4.2% true collision rate (i.e., collision without detection) while NV had only 0.6% (relative risk increase = 6.7). 70% of true collisions in HH occurred on the blind side even though their head scanning was biased toward their blind side (F(1,68) = 5.33, p = 0.02) during walking. All other collisions in HH and NV were on the 60° seeing side but out of the HMD’s FoV. However, detection rates (F(1,12) = 2.84, p = 0.12) and response times (F(1,12) = 2.08, p = 0.18) were not different between two groups.

Conclusions : We developed the realistic collision test on HMDs with actual walking and gaze scanning. In this realistic walking, the traditional detection rate and response time were not different between HH and NV due to the head scanning in HH, but the collision rate showed practical difficulty in HH. This may suggest that collision avoidance could be a possible objective outcome for mobility performance.

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

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×