Measuring attentional bias to peripheral facial deformities

Abstract

Objectives: Introduce a novel method for objectively evaluating attentional bias to peripheral facial deformities using an established metric of attention.

Methods: The SMI eye-tracker system (SensoMotoric Inc., Boston, MA) was used to record the eye movement patterns, called scanpaths, of eight naïve observers gazing at pictures of faces with or without peripheral surgical deformities. The scanpaths of observers gazing on those novel faces were compared, and the fixation durations for different facial regions were compared between faces.

Results: There were statistically significant differences in the mean fixation times between the faces considered normal and those considered abnormal (those with an obvious defect). When multivariate analysis of variance was performed with dependent variables total fixation time, fixation time in central triangle, and fixation time in the defect region and the independent variable face, all four tests were highly statistically significant. When univariate analysis of variance was performed to test the hypothesis that defect fixation times varied by face, the results were highly statistically significant (F = 8.79, P = .0003).

Conclusions: Observers gazing on faces typically focus their attention on discriminating features, such as eyes, nose, and mouth. The well-established method of eye movement recordings was applied in a novel way to provide quantitative data showing changes in observer gaze patterns to focus on deformities. These gaze patterns are a direct reflection of observer attention. This is the first objective method to quantify the amount of distraction caused by peripheral facial deformities and may provide insight into the perception of facial deformity.