Detecting nursing students' empathy in video-recorded simulation using computer vision approach

Objective: To evaluate the feasibility of artificial intelligence (AI) algorithms in assessing empathy demonstrated by nursing students during simulation-based communication scenarios. The research compared traditional empathy assessments with algorithm-based analyses to explore the potential of innovative methods for standardizing the measurement of empathy in clinical and educational settings. Background: Empathy is a cornerstone of the healthcare and nursing professions; however, it remains challenging to measure effectively. Design: An observational comparative study. Methods: The study was conducted with a sample of 37 nursing students in their second to fourth year of studies. Empathy was measured through traditional tools completed by observers to assess general sense of empathy and the extent of using six empathic body gestures. In addition, video-based analyses with AI algorithms assessed participants' empathic body gestures of eye contact, smiling, physical contact, and closeness. Data analyses included descriptive and correlations between traditional and innovative approaches. Results: Body gestures were positively associated with Sense of empathy (correlation coefficients ranging from 0.41 to 0.63). Moderate positive correlations were found between the physical contact detection algorithm and four observer-reported measures (0.46 – 0.54, p < .01). Weak negative correlations were found between the smile algorithm and three observer-reported measures (-0.37 – -0.33, p < .05). Conclusions: AI-driven technologies offer an effective approach to evaluating communication skills within health education processes. Integrating innovative methods has the potential to streamline training, reduce costs, and enhance students' communication competencies, which in turn increases patient satisfaction. Further research is needed to refine and validate the proposed methods to ensure greater accuracy and effectiveness.