Skip to main content
. 2021 Jan 1;11(5):2098–2107. doi: 10.7150/thno.48027

Figure 1.

Figure 1

The structure of deep learning network. The entire network includes a feature extraction module (a convolutional network) for extracting deep learning features, and a classification module (a fully-connected network) for classifying PD-L1 expressions based on deep learning features combined with predefined radiomic features and clinical characteristics.