Fig. 2
From: DeepHipp: accurate segmentation of hippocampus using 3D dense-block based on attention mechanism
The overall architecture of the DeepHipp model. The model contains an encoding part and a decoding part. The encoder is on the left side of the auxiliary line, and the decoder is on the right side. In the encoding section, each module consists of a dense block. In the decoding section, each module consists of an attention block. The model uses end-to-end input and output to segment the hippocampus from the whole brain