Extended Data Fig. 2: Detection accuracy per condition on the IND1 dataset. | Nature Medicine

Extended Data Fig. 2: Detection accuracy per condition on the IND1 dataset.

From: Collaboration between clinicians and vision–language models in radiology report generation

Extended Data Fig. 2

The receiver operating characteristic (ROC) curve of the Flamingo-CXR report generation model, shown along with the true positive rate (TPR) and false positive rate (FPR) pairs for two certified radiologists are shown for 6 conditions for which the expert labels were collected. The operating point of our model with the default inference scheme (Beam 3) is also shown. Error bars represent 95% confidence intervals (calculated using bootstrapping with 1000 repetitions).

Back to article page