Skip to main content
. 2024 Oct 4;13(10):20584601241288509. doi: 10.1177/20584601241288509

Figure 2.

Figure 2.

Convolutional neural network model. The center slice and pre- and post-slices from CT images of thymomas in the axial direction were used as three channel inputs, and training was performed using a 2-dimensional convolutional neural network, which was a fine-tuning model based on VGG16. CT, computed tomography.