Fig. 6From: Automatic literature screening using the PAJO deep-learning model for clinical practice guidelinesBoxplots of article applicability + quality. The y-axis reflects the interquartile model predictions, where the red dots are outliers. The x-axis reflects the scale of expert-provided ground-truth applicability + quality, with “8” being the highest. The left panel reflects articles published in 2022, and the right panel reflects articles published in 2021Back to article page