文献类型：期刊文献

英文题名：Dilated Dense U-Net for Infant Hippocampus Subfield Segmentation

作者：Zhu, Hancan[1,2,3];Shi, Feng[4];Wang, Li[1,2];Hung, Sheng-Che[1,2];Chen, Meng-Hsiang[5];Wang, Shuai[1,2];Lin, Weili[1,2];Shen, Dinggang[1,2,6]

机构：[1]Univ North Carolina Chapel Hill, Dept Radiol, Chapel Hill, NC 27599 USA;[2]Univ North Carolina Chapel Hill, Biomed Res Imaging Ctr, Chapel Hill, NC 27599 USA;[3]Shaoxing Univ, Sch Math Phys & Informat, Shaoxing, Peoples R China;[4]Shanghai United Imaging Intelligence Co Ltd, Shanghai, Peoples R China;[5]Chang Gung Univ, Coll Med, Kaohsiung Chang Gung Mem Hosp, Dept Diagnost Radiol, Kaohsiung, Taiwan;[6]Korea Univ, Dept Brain & Cognit Engn, Seoul, South Korea

年份：2019

卷号：13

外文期刊名：FRONTIERS IN NEUROINFORMATICS

收录：SCI-EXPANDED(收录号：WOS:000465702800001)、、Scopus(收录号：2-s2.0-85067412559)、WOS

基金：This work utilizes data collected by a NIH grant (1U01MH110274) and the efforts of the UNC/UMN Baby Connectome Project Consortium. HZ was partially supported by National Natural Science Foundation of China (Nos. 61602307, 61877039), and Natural Science Foundation of Zhejiang Province (No. LY19F020013).

语种：英文

外文关键词：fully convolutional network; dilated dense network; deep learning; hippocampal subfield segmentation; infant hippocampus

外文摘要：Accurate and automatic segmentation of infant hippocampal subfields from magnetic resonance (MR) images is an important step for studying memory related infant neurological diseases. However, existing hippocampal subfield segmentation methods were generally designed based on adult subjects, and would compromise performance when applied to infant subjects due to insufficient tissue contrast and fast changing structural patterns of early hippocampal development. In this paper, we propose a new fully convolutional network (FCN) for infant hippocampal subfield segmentation by embedding the dilated dense network in the U-net, namely DUnet. The embedded dilated dense network can generate multi-scale features while keeping high spatial resolution, which is useful in fusing the low-level features in the contracting path with the high-level features in the expanding path. To further improve the performance, we group every pair of convolutional layers with one residual connection in the DUnet, and obtain the Residual DUnet (ResDUnet). Experimental results show that our proposed DUnet and ResDUnet improve the average Dice coefficient by 2.1 and 2.5% for infant hippocampal subfield segmentation, respectively, when compared with the classic 3D U-net. The results also demonstrate that our methods outperform other state-of-the-art methods.