Eye-tracking analyses of physician face gaze patterns in consultations

Area of Science:

• Behavioral Science
• Human-Computer Interaction
• Medical Communication

Background:

• Face gaze is a fundamental non-verbal behavior, essential for social interaction and communication.
• Current methodologies for assessing face gaze using eye-tracking glasses lack standardized guidelines for selecting appropriate measures.
• Existing face gaze analysis methods are often time-consuming and require significant manual effort.

Purpose of the Study:

• To compare the utility of duration, frequency, and dwell time as measures of face gaze.
• To determine if early face gaze patterns predict overall interaction face gaze patterns.
• To provide methodological guidance for researchers studying face gaze in clinical settings.

Main Methods:

• Secondary analysis of mobile eye-tracking data from 16 physicians and 100 patient consultations.
• Comparison of three face gaze measures: duration, frequency, and dwell time.
• Assessment of predictive validity of face gaze patterns within the first 30, 60, and 120 seconds of consultations.

Main Results:

• Duration and frequency of face gaze were found to be unrelated, highlighting potential variability in research findings based on the chosen measure.
• Dwell time demonstrated a correlation with both duration and frequency of face gaze.
• Face gaze patterns observed in the initial seconds of consultations significantly predicted patterns throughout the entire interaction (R² = 0.26 to 0.73).

Conclusions:

• Early face gaze behavior in medical consultations can reliably predict overall gaze patterns.
• Dwell time emerges as a potentially more comprehensive measure, correlating with both duration and frequency.
• These findings offer valuable insights for optimizing methodological choices in future face gaze research.