Using AI to Detect Errors in Medical Notes
Medical notes are vital for patient care, but even small mistakes can have serious consequences. This work explores whether AI language models can help identify potential errors in clinical text. We find that better instructions can make these systems much more effective, raising important questions about how AI might support safer healthcare in the future.
Keywords
health informatics, artificial intelligence, patient safety, healthcare, medical notes, language models, natural language processing, clinical text
Staff
Craig Myles, Patrick Schrempf, David Harris-Birtill