Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

Illustration of the systematic workflow employed in this study.

More »

Fig 1 Expand

Fig 2.

The LC25000 dataset where (a) adenocarcinoma, (b) benign, and (c) squamous cell carcinoma.

More »

Fig 2 Expand

Fig 3.

The proposed XLLC-Net architecture.

More »

Fig 3 Expand

Table 1.

Summary of the proposed model, detailing the number of parameters and characteristics of each layer.

More »

Table 1 Expand

Table 2.

Hyper-parameters employed for training XLLC-Net.

More »

Table 2 Expand

Table 3.

Specifications of the system used for the proposed framework.

More »

Table 3 Expand

Table 4.

Performance metrics (Accuracy, Precision, Recall, and F1-score) for five independent training trials of the XLLC-Net model.

More »

Table 4 Expand

Table 5.

Performance comparison of different DL models in predicting cancer types.

More »

Table 5 Expand

Fig 4.

Performance of the proposed model against various DL models.

More »

Fig 4 Expand

Table 6.

Performance of the proposed model against various DL models.

More »

Table 6 Expand

Table 7.

State-of-the-art comparison between proposed model and previous models on lung histopathological data.

More »

Table 7 Expand

Fig 5.

Comparison of training and validation accuracy across epochs for several pre-trained models, including (a) AlexNet, (b) ResNet50, (c) VGG16, and (d) VGG19.

More »

Fig 5 Expand

Fig 6.

Training and validation accuracy across epochs for our proposed XLLC-Net model. (a) 1st Trial, (b) 2nd Trial, (c) 3rd Trial, (d) 4th Trial, (e) 5th Trial.

More »

Fig 6 Expand

Fig 7.

Comparison of training and validation loss across epochs for several pre-trained models, including (a) AlexNet, (b) ResNet50, (c) VGG16, and (d) VGG19.

More »

Fig 7 Expand

Fig 8.

Training and validation loss across epochs for our proposed XLLC-Net model. (a) 1st Trial, (b) 2nd Trial, (c) 3rd Trial, (d) 4th Trial, (e) 5th Trial.

More »

Fig 8 Expand

Fig 9.

Comparison of AUC for various pre-trained models, including (a) AlexNet, (b) ResNet50, (c) VGG16, and (d) VGG19.

More »

Fig 9 Expand

Fig 10.

AUC for our proposed XLLC-Net model across five independent trials. (a) 1st Trial, (b) 2nd Trial, (c) 3rd Trial, (d) 4th Trial, (e) 5th Trial.

More »

Fig 10 Expand

Fig 11.

Comparison of confusion matrix for several pre-trained models, including (a) AlexNet, (b) ResNet50, (c) VGG16, and (d) VGG19.

More »

Fig 11 Expand

Fig 12.

Confusion matrices of the XLLC-Net model across five independent trials. (a) 1st Trial, (b) 2nd Trial, (c) 3rd Trial, (d) 4th Trial, (e) 5th Trial.

More »

Fig 12 Expand

Fig 13.

Saliency map of squamous cell carcinoma.

More »

Fig 13 Expand

Fig 14.

Saliency map of benign.

More »

Fig 14 Expand

Fig 15.

Saliency map of adenocarcinoma.

More »

Fig 15 Expand

Fig 16.

GradCAM of squamous cell carcinoma.

More »

Fig 16 Expand

Fig 17.

GradCAM of benign.

More »

Fig 17 Expand

Fig 18.

GradCAM of adenocarcinoma.

More »

Fig 18 Expand