Fig. 2
From: Deep learning based dual stage model for accurate nasogastric tube positioning in chest radiographs
Visualization results of prediction masks and class activation map (Grad-CAM) on external validation datasets. Prediction masks are extracted from nnU-Net and Grad-CAM results are from MedCLIP pre-trained ResNet-50. (a) X-ray images, (b) label of the segmentation model, (c) results with the sum of (a) and prediction of the segmentation, (d) Grad-CAM results. The top two low are the incorrectly predicted cases(incomplete, complete) and the bottom two low(incomplete, complete) are the correctly predicted cases.
