Using AI to Strengthen Patient History-Taking Skills across English and Mandarin for Future Clinicians
This proposal outlines the development of an AI-powered tool to enhance medical students' patient history-taking skills across English and Mandarin. The objective is to better prepare students for the clinical environment by improving their listening comprehension and clinical competencies in patient history-taking, prior to hospital-based rotations. The tool will be built upon a large language model, Apollo-72B, developed by the Shenzhen Research Institute of Big Data. By fine-tuning this model, the tool will generate diverse patient complaint scenarios for students to practice. Presented with an audio recording, students will document a structured patient history and compare their work to an example. The effectiveness will be evaluated by tracking practice sessions and assessing students' history-taking abilities during clinical encounters. The expectation is that students who use the tool more frequently will demonstrate stronger patient history-gathering skills in real-world settings, indicating the tool's value in developing this crucial competency before clinical rotations. This AI-powered practice tool aims to create a safe, controlled environment for students to hone their patient intake and information-gathering skills across languages, preparing them for successful clinical interactions during training and future practice.