AI-human collaborative approaches in emotional training: applying visual thinking strategies as an analytical framework for medical students’ drawings from art-based activities

Abstract Purpose This study examines AI-human collaborative emotion training for medical students using an art-based intervention with visual thinking strategies (VTS) and metacognitive reflection. It aims to measure pre-post changes in emotion appraisal, empathy, and well-being, along with post-intervention reflective writings. Methods A quasi-experimental design compared medical students from an elective course in 2023 (control, n = 35) and 2024 (intervention, n = 33), for a total of 68 participants. The six-week intervention consisted of art-based activities three hours per week. AI-human collaborative emotion training used VTS as an interpretive framework where participants uploaded digital artwork to AI platforms, corrected AI interpretations, and compared these with peer feedback and their original intent, fostering metacognitive engagement in reflecting on the recognition and perception of both one’s own and others’ emotions. Outcome measures were emotion appraisals recorded using survey and reflective writings. Results Self-emotion appraisal significantly improved in the intervention group (pre: 5.13; post: 5.40; p = 0.035) but not in the control group (p = 0.501). The significant finding remained significant after controlling for age and gender (p = 0.027). Others’ emotion appraisal showed no significant changes. Qualitative analysis revealed AI excelled in precise visual details and symbolism, while peers provided holistic, experience-based interpretations. Conclusions AI-assisted VTS art analysis, combined with metacognitive reflection, enhanced emotional appraisal, showing how technology and human interaction complement emotional learning, offering a balanced framework for innovation and human engagement in medical education.