Fig. 4: Three examples of the YOLOX model for the identification and localization of multiple abnormalities in CXRs. | Nature Communications

Fig. 4: Three examples of the YOLOX model for the identification and localization of multiple abnormalities in CXRs.

From: A deep-learning-based framework for identifying and localizing multiple abnormalities and assessing cardiomegaly in chest X-ray

Fig. 4

a In the case A, the category and localization of bounding boxes predicted by the YOLOX model were highly coincided with the GT bounding boxes; b In the case B, the YOLOX model missed atelectasis in the right lung hilar region and merged two separate calcifications on the aortic arch into a single calcification; c In the case C, The YOLOX model missed microcalcifications on the aortic arch and additionally diagnosed consolidation around the right lung mass. However, the other bounding boxes predicted by YOLOX model in case B and C were also highly coincided with the GT bounding boxes. GT: ground-truth.

Back to article page