Skip to main content
. 2023 Aug 3;4(9):100802. doi: 10.1016/j.patter.2023.100802

Table 2.

Metric scores of findings-generating models, including M2 Trans, and the random retrieval baseline model

BLEU BERTScore CheXbert vector similarity RadGraph F1
M2 Trans 0.220∗ 0.386∗ 0.452∗ 0.244∗
Random retrieval of findings 0.123 0.323 0.235 0.105

The 95% confidence interval and range of metric scores are available in Table S6.

∗

indicates the best-performing model.