Towards Conversational Diagnostic AI

Towards Conversational Diagnostic AI

11 Jan 2024 | Tao Tu*, Anil Palepu*, Mike Schaekermann*, Khaled Saab, Jan Freyberg, Ryutaro Tanno, Amy Wang, Brenna Li, Mohamed Amin, Nenad Tomasev, Shekoofeh Azizi, Karan Singhal, Yong Cheng, Le Hou, Albert Webson, Kavita Kulkarni, S. Sara Mahdavi, Christopher Senturs, Juraj Gottweis, Joelle Barral, Katherine Chou, Greg S. Corrado, Yossi Matias, Alan Karthikesalingam†, Vivek Natarajan†
The paper introduces AMIE (Articulate Medical Intelligence Explorer), a Large Language Model (LLM) designed for diagnostic dialogue in medicine. AMIE is optimized for clinical history-taking and diagnostic reasoning through a self-play based simulated environment with automated feedback mechanisms. The authors designed a framework to evaluate AMIE's performance on clinically meaningful axes, including history-taking, diagnostic accuracy, management reasoning, communication skills, and empathy. A randomized, double-blind crossover study was conducted using text-based consultations with validated patient actors, comparing AMIE to primary care physicians (PCPs). AMIE demonstrated superior performance on 28 out of 32 axes according to specialist physicians and 24 out of 26 axes according to patient actors. The study highlights the potential of conversational AI in improving accessibility, consistency, and quality of medical care. However, the research has limitations, such as the use of a text-chat interface, which may not fully represent real-world clinical practice. Further research is needed to translate AMIE to real-world settings.The paper introduces AMIE (Articulate Medical Intelligence Explorer), a Large Language Model (LLM) designed for diagnostic dialogue in medicine. AMIE is optimized for clinical history-taking and diagnostic reasoning through a self-play based simulated environment with automated feedback mechanisms. The authors designed a framework to evaluate AMIE's performance on clinically meaningful axes, including history-taking, diagnostic accuracy, management reasoning, communication skills, and empathy. A randomized, double-blind crossover study was conducted using text-based consultations with validated patient actors, comparing AMIE to primary care physicians (PCPs). AMIE demonstrated superior performance on 28 out of 32 axes according to specialist physicians and 24 out of 26 axes according to patient actors. The study highlights the potential of conversational AI in improving accessibility, consistency, and quality of medical care. However, the research has limitations, such as the use of a text-chat interface, which may not fully represent real-world clinical practice. Further research is needed to translate AMIE to real-world settings.
Reach us at info@study.space