Fig. 5: Specialty prediction accuracy: matched vs. at least one specialty match evaluation. | npj Digital Medicine

Fig. 5: Specialty prediction accuracy: matched vs. at least one specialty match evaluation.

From: Evaluating large language model workflows in clinical decision support for triage and referral and diagnosis

Fig. 5

This figure compares the model performance in predicting medical specialties. The graph shows the accuracy [%] for two evaluation methods: matched specialty (left) and atleast one specialty match (right). Results are displayed for both evaluation types, illustrating the differences in model accuracy across the different LLMs for specialty prediction tasks.

Back to article page