Fig. 2From: Reducing pediatric total-body PET/CT imaging scan time with multimodal artificial intelligence technologySchematic diagram of the 3D convolutional neural network (p3DNet) used in this work. Two modal-specific encoders and one decoder that synthesizes the full-dose PET images are included. The arrows indicate the flow of computational operations, and the number of input and output feature images for each module is marked below its boxBack to article page