AI-powered platform revolutionizing blood cell morphology education for medical students

Abstract

Background: This study aims to preliminarily explore the advantages and potential issues of artificial intelligence in the teaching of blood cell morphology to undergraduate medical students, so as to provide theoretical support and practical experience for promoting the intelligent transformation of medical education. Methods: Undergraduate students from the 2021 cohort of the Aerospace School of Clinical Medicine at Peking University were assigned as the experimental group, while students from the 2020 cohort served as the control group. The experimental group utilized the AI platform to study blood cell morphology, whereas the control group relied on conventional teaching methods. We compared the accuracy rates of cell identification between two groups of students. Additionally, we conducted supplementary research through questionnaires, post-class interviews, and classroom observations. Results: The experimental group achieved a significantly higher average score in cell identification (87.82 ± 9.63) compared to the control group (74.83 ± 12.41) (P<0.0001). The correct identification rates of metamyelocytes, eosinophils, and monocytes in the experimental group were significantly increased by over 30%. Discussion: AI holds considerable promise in medical education, particularly in the instruction of hematology cell morphology. Nevertheless, further research is required. Traditional microscope-based teaching should not be completely dismissed, as current digital platforms for blood cells do not yet capture all cellular nuances.