Shaping learning objectives for biomedical artificial intelligence: Student-centered insights into novel cell visualization technology.

Background: Artificial intelligence (AI) technology is rapidly entering biomedical research, and there is a need to assess and develop curricula that address trainees' learning objectives and interests. Studies of biomedical workforce development show that the prospective engagement of students in formulating educational objectives and activities improves motivation and learning outcomes. This study aimed to explore the educational applications of a novel AI-powered technology in undergraduate education. Methods: A mixed-methods approach using elicitation interviews and cultural domain analysis was applied to identify the salience of ideas around the educational uses of Functional Unit State Identification & Navigation with Whole Slide Images (FUSION), an AI-powered cell-visualization technology. Interviews from 21 students were reduced to learning application statements and assessed for cultural salience and clustering for potential educational applications. Results: Saturation was reached after 11 interviews, and analysis resulted in eight clusters of 25 unique consensus-based statements. Students thought of the technology as a tool for cell analysis and measuring, but they also viewed applications for medical and K-12 education, public engagement, and note-keeping for technology research. Methodologically, our study demonstrates the potential of cultural consensus for learner-centered curriculum development. Conclusions: Our findings suggest that trainees perceive many educational uses of FUSION, including those that fit traditional biomedical research curricula and translational applications. Trainees should be engaged in co-design to support and guide technology translation for educational use.