Normal Gaze Patterns of the Face in Lateral View

Importance: Current efforts to quantify the attentional distraction of facial deformities have been limited to deformities that are best perceived when the face is in frontal view, and there remains a paucity of knowledge of societal perception of the face in lateral view. To date, no attempts have been made to characterize the fixation patterns of the face in lateral view. Objective: To characterize the fixation patterns and gaze patterns of the face in lateral view. Design, Setting, and Participants: This was a prospective randomized controlled trial at an academic tertiary medical center. Eighty participants (mean age 23.6 ± 1.7 years, 52.5% female) gazed freely at 11 images of faces in lateral view for 10 s each as an infrared eye-tracker recorded eye movements in real time. Main Outcomes and Measures: Recorded eye movements are superimposed on the images to visualize areas of interest (AOIs) that attract the most attention. Fixation duration targeted at each AOI is transformed into relative fixation of the entire face and neck. Hotelling's test of variance followed by post hoc t-testing assessed for the significance of fixation differences between the mouth and cheeks. Results: Participants spent an average of 6.0 (95% CI 5.8-6.2) s gazing at the face and neck areas of each image. Of this attention, 2.8 s or 49.8% (45.4-54.2%) was directed toward the eye, followed distantly by the nose (mean, 95% CI) (16.4%, 14.5-18.3%), cheek (12.0%, 11.4-14.5%), neck (4.6%, 3.8-5.4%), and mouth (4.0%, 3.4-4.6%). These differences were found to be significant by Hotelling's analysis and post hoc testing. A student's t-test also indicated observers directed significantly more attention within the eye-nose-cheek triangle [4.6 s or 79.5% (75.6-83.5)] compared with the eye-nose-mouth triangle [4.1 s or 71.2% (66.9-75.5)] (p < 0.001). Conclusions and Relevance: When perceiving novel faces in lateral view, casual observers preferentially directed attention toward the eye, nose, and cheek. These findings suggest that we draw from a slightly different collection of features to build a schema of the sagittal face, which may serve to complement the central triangle and build upon a three-dimensional model of the "normal" human face. Level of Evidence: NA.