Perceptions of 1st year undergraduate medical students on artificial intelligence: mixed-methods survey study from a rural medical college of West Bengal

Abstract

Background: Artificial Intelligence (AI) is increasingly influencing healthcare practices yet remains under-represented in undergraduate medical curricula. Understanding student perceptions of AI is crucial for informed curriculum development, particularly in rural medical settings. Objectives: To assess the AI literacy, attitudes, and perceptions among first-year MBBS students using the Meta-Artificial Intelligence Literacy Scale (MAILS), and to explore gender difference of AI literacy, perceived advantages and disadvantages of AI integration in medical education. Methods: This descriptive, cross-sectional, mixed-methods study was conducted among 88 first-year MBBS students at Rampurhat Government Medical College, West Bengal. Quantitative data were collected using the validated MAILS questionnaire, encompassing 34 items across 8 domains. Qualitative responses were obtained from two open-ended questions on AI's perceived pros and cons. Data analysis included independent t-tests and thematic analysis using NVivo software. Results: Students demonstrated highest proficiency in Persuasion Literacy (57.73%) and Use & Apply AI (56.5%). Lowest scores were in Create AI (37.6%) and Learning (49.73%), indicating gaps in technical and adaptive skills. No statistically significant gender differences were observed (p = 0.39). Thematic analysis revealed students value AI for simplifying tasks and enhancing diagnostics but expressed concerns over loss of empathy, dependency, and privacy breaches. Control, emotion, and interaction were the most developed competencies.showed no statistically significant difference in AI literacy between genders. Conclusion: Students are confident users but not creators of AI. Their moderate AI literacy and mature ethical concerns highlight the urgent need for structured, ethically grounded AI training in medical education, especially in resource-limited rural contexts.

Ähnliche Arbeiten