Purchase this article with an account.
U. Schiefer, G. Hardiess, F. Schaeffel, H. Wiethoelter, H.O. Karnath, R. Vonthein, B. Schoenfisch, H.A. Mallot, E. Papageorgiou; To What Extent Can Visual Exploration Compensate for Homonymous Scotomas? A Pilot Study Based on Virtual Reality Driving Tasks . Invest. Ophthalmol. Vis. Sci. 2006;47(13):792.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess the effect of exploratory compensation in patients with homonymous visual field defects (HVFDs) on the frequency of traffic accidents under standardised virtual reality (VR) conditions.
Size and location of absolute HVFDs were assessed by binocular semi–automated kinetic perimetry (stimulus III4e [26', 320 cd/m²], angular velocity 3°/s, background luminance 10 cd/m²) with the OCTOPUS 101 perimeter (HAAG–STREIT Inc., Koeniz, Switzerland). Perimetric results were digitally superimposed onto the standardised VR driving tasks (30 sequences of approaching an intersection with two levels of cross–traffic density). For analysis, each of the following numbers of cars (a – d) was normalised by the number of video frames of each scene with unchanged conditions: a: within intact visual field (VF) AND gaze field (GF), b: within affected VF but intact GF, c: within intact VF but affected GF, d: within affected VF AND GF.
Perimetric information intake was defined as PI = (a + c) / n, and explorative information intake as EI = (a + b) / n, with n being the total number of cars. Efficiency of exploration was defined as EoE = b / (b + c). Linear regressions from the number of accidents by PI and EI were estimated. Also the ratio EI/PI versus EoE was assessed.
9 (4 females, 5 males; age range 21 to 70 years) patients with HVFDs due to unilateral vascular or traumatic brain lesions with a minimum visual acuity of 10/20 participated in this study.
There was a strong relationship between EI/PI ratio and EoE (R^2 = 0.78, linear regression model), i.e. patients with high efficiency of exploration are characterized by an improvement of information intake via exploratory eye and head movements. EI (R^2 = 0.45, linear regression model) is able to predict the number of accidents substantially better than PI (R^2 = 0.02, linear regression model). The mean values of PI und EI were 0.57 and 0.63 respectively.
Perimetric findings per se seem to be inadequate in predicting driving performance of patients with homonymous VFDs under VR conditions. Compensation via exploratory eye and head movements has to be considered additionally.
This PDF is available to Subscribers Only