Skip to main content
Fig. 5 | BMC Medical Imaging

Fig. 5

From: Transfer learning–based PET/CT three-dimensional convolutional neural network fusion of image and clinical information for prediction of EGFR mutation in lung adenocarcinoma

Fig. 5

TS_TL predicted tumor-associated areas for subsolid lesions with either EGFR wild-type or mutation. For each submap, the input CT or PET image, the attention map, and the model-predicted tumor-associated areas are from left to right. For LADC tumors, the deep learning model generated an attention map indicating the importance of each part of the tumor; high-reaction regions (predicted tumor-associated areas) were retained with a cutoff value of 0.5. P− and P+ represented the predicted probability of EGFR wild-type and mutant, respectively

Back to article page

BMC Medical Imaging

ISSN: 1471-2342

Contact us