Artificial Intelligence in Medicine

The Lab’s work on artificial intelligence in medicine centers on the responsible integration of AI into assessment and learning, with a focus on supporting—not replacing—human judgment, reflection, and professional growth. Rather than treating AI as a predictive or automation tool alone, our research frames AI as a co-collaborative system that can augment clinical reasoning assessment, narrative feedback, and educational decision-making when grounded in psychometric theory, learning science, and qualitative inquiry. Through platforms such as Night-onCall and FeedbackAssist, we study how AI can generate timely, interpretable, and equitable feedback, surface patterns across complex performance data, and make assessment processes more transparent to learners and educators. A central emphasis of this work is validity, trust, and

consequences: examining how AI systems perform across contexts and populations, how their outputs are interpreted and acted upon, and how design choices influence equity, professional identity formation, and the lived experience of medical training. Together, this research advances a human-centered model of AI in medicine that aligns technological innovation with educational values, ethical responsibility, and clinical competence.