Changes in head and eye movements during navigation with peripheral visual field loss

Poster Presentation 53.337: Tuesday, May 19, 2026, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Scene Perception: Virtual reality

Sam Beech1 (), Maggie McCracken1, Bobby Bodenheimer2, Sarah Creem-Regehr1; 1University of Utah, 2Vanderbilt University

Simulated and glaucoma-related peripheral visual field loss (PVFL) impairs navigation. Effective navigators optimally integrate visual cues with body-based self-motion cues according to their relative reliability. PVFL may reduce the spatial reliability of vision, but compensatory visual search behaviors could preserve reliability. To investigate this, we simulated gaze-contingent PVFL during a virtual reality navigation task while recording head and eye movements. We present preliminary data for 15 healthy adults across three visual field groups: 110°, 30°, or 20°. Participants walked to three markers with landmarks in the background. The markers were removed, and participants walked to their estimated position of the first marker (homing task) in three conditions: Full Cues – no manipulation; Self-Motion Only – landmarks removed; or Vision Only – participants rotated to disrupt directional self-motion cues. We recorded error and variability. Across groups, error and variability were greater in single-cue conditions: Full Cues (110° = 59 (40) cm, 30° = 55 (37) cm, 20° = 56 (31) cm); Vision Only (110° = 76 (46) cm, 30° = 54 (37) cm, 20° = 77 (53) cm); and Self-Motion Only (110° = 92 (49) cm, 30° = 89 (48) cm, 20° = 94 (52) cm), d ≥ 0.411/0.459. In the Vision Only condition, average absolute yaw-plane head-trunk deviation was greater with 30° (25.0°) compared to 110° (16.0°) and 20° (18.5°) vision, d ≥ 0.39. Additionally, head and eye pitch were lower with 20° (-19.4°/-37.5°) and 30° (-14.5°/-34.2°) compared to 110° (-7.9°/-22.0°) vision, d ≥ 0.66/1.00). The lack of an effect of PVFL on navigation accuracy and variability could be explained by compensatory visual search. The 30° and 20° groups showed increased fixation on landmark-ground intersections, and the 30° group exhibited increased side-to-side head movements between landmarks. These results show preliminary support for distinct visual search behaviors during navigation with PVFL.