Fully Automated Delineation of Gross Tumor Volume for Head and Neck Cancer on PET-CT Using Deep Learning: A Dual-Center Study

Figure 1

Architecture of the proposed CNN model. The proposed network includes two phases: feature representation phase and scores map reconstruction phase. The feature representation phase is composed of 5 downsampling blocks (conv-ReLU-pool layer), 4 convolution layers, and 4 ReLU layers. The scores map reconstruction phase consists of 5 upsampling blocks (deconv-concat-conv-ReLU layer), a convolution layer, and an ReLU layer. K, filter size; P, zero padding; S, stride.