Towards Accurate Fetal Brain Parcellation via Hierarchical Network and Loss

doi:10.1007/978-3-031-73260-7_7

	Towards Accurate Fetal Brain Parcellation via Hierarchical Network and Loss
	Huang, Shijie1 ; Zhang, Kai1 ; Huang, Jiawei1,2 ; Kong, Lingnan 3; Zhu, Fangmei 4; Ding, Zhongxiang 4; Chen, Geng 5; Shen, Dinggang1,6,7
	2025
会议录名称	LECTURE NOTES IN COMPUTER SCIENCE (INCLUDING SUBSERIES LECTURE NOTES IN ARTIFICIAL INTELLIGENCE AND LECTURE NOTES IN BIOINFORMATICS)
ISSN	0302-9743
卷号	14747 LNCS
页码	70-81
发表状态	已发表
DOI	10.1007/978-3-031-73260-7_7
摘要	Automatic fetal brain parcellation on Magnetic Resonance (MR) images is increasingly being used to assess prenatal brain growth and development. Despite their progress, existing methods are limited due to ignoring of the hierarchical nature of segmentation labels and the rich complementary information among hierarchical labels. To address these limitations, we propose a novel deep-learning model to segment the whole fetal brain into 87 fine-grained regions hierarchically. Specifically, we design a hierarchical network with adjustable levels and define a three-level structure. These levels are dedicated, respectively, to predicting 8 types of brain tissues, 36 more detailed brain regions, and ultimately 87 brain regions according to developing Human Connectome Project (dHCP) labels. The coarse-level network is capable of providing prior features to the fine-level network for fine-grained brain parcellation. This design involves decomposing complex problems into simpler ones and addresses intricate issues with the priors for resolving simple problems. Furthermore, we design a data augmentation module to simulate variations in scanning parameters, enabling precise segmentation of fetal brain images across diverse domains. Finally, we integrate this data augmentation module into a semi-supervised paradigm to alleviate the shortage of high-quality labeled data and enhance the generalizability of our model. Thanks to these designs, our model can obtain fine-grained and multi-scale brain segmentation with high robustness to variations in MR scanners and imaging protocols. Extensive experiments on 558 dHCP and 176 fetal brain MR images demonstrate that our model achieves state-of-the-art segmentation performance across multi-site datasets. Our code is publicly available at https://github.com/sj-huang/HieraParceNet. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
关键词	Brain mapping Deep learning Image coding Image enhancement Image segmentation Labeled data Neonatal monitoring Nuclear magnetic resonance Self-supervised learning Semi-supervised learning Brain regions Data augmentation Fetal brain Fine grained Hierarchical model Hierarchical network Human Connectome Segmentation Semi-supervised learning Simple++
会议名称	9th International Workshop on Perinatal, Preterm and Paediatric Image Analysis, PIPPI 2024, held in Conjunction with the 27th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2024
出版地	GEWERBESTRASSE 11, CHAM, CH-6330, SWITZERLAND
会议地点	Marrakesh, Morocco
会议日期	October 6, 2024 - October 6, 2024
URL	查看原文
收录类别	EI ; CPCI-S
语种	英语
资助项目	National Natural Science Foundation of China["62131015","62250710165","62203355","U23A20295","2022ZD0209000"] ; Shanghai Municipal Central Guided Local Science and Technology Development Fund[YDZX20233100001001] ; Key R&D Program of Guangdong Province, China["2023B0303040001","2021B0101420006"]
WOS研究方向	Computer Science ; Neurosciences & Neurology ; Obstetrics & Gynecology ; Radiology, Nuclear Medicine & Medical Imaging
WOS类目	Computer Science, Artificial Intelligence ; Neuroimaging ; Obstetrics & Gynecology ; Radiology, Nuclear Medicine & Medical Imaging
WOS记录号	WOS:001437147800007
出版者	Springer Science and Business Media Deutschland GmbH
EI入藏号	20244417277454
EI主题词	Magnetic resonance imaging
EISSN	1611-3349
EI分类号	101.1 ; 1101.2 ; 1101.2.1 ; 1106.3.1 ; 1301.2.2 ; 709 Electrical Engineering, General ; 746 Imaging Techniques ; 903.1 Information Sources and Analysis
原始文献类型	Conference article (CA)
文献类型	会议论文
条目标识符	https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/449159
专题	信息科学与技术学院_博士生生物医学工程学院_PI研究组_沈定刚组
通讯作者	Chen, Geng; Shen, Dinggang
作者单位	1.School of Biomedical Engineering and State Key Laboratory of Advanced Medical Materials and Devices, ShanghaiTech University, Shanghai; 201210, China 2.Department of Orthopedic Surgery, the First Affiliated Hospital, Zhejiang University, Hangzhou; 310003, China 3.Department of Radiology, the First Affiliated Hospital, Nanjing Medical University, Nanjing; 210000, China 4.Department of Radiology, the Affiliated Hangzhou First People’s Hospital, Westlake University, Hangzhou; 310006, China 5.National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, School of Computer Science and Engineering, Northwestern Polytechnical University, Xi’an; 710072, China 6.Shanghai United Imaging Intelligence Co., Ltd., Shanghai; 200230, China 7.Shanghai Clinical Research and Trial Center, Shanghai; 201210, China
第一作者单位	上海科技大学
通讯作者单位	上海科技大学
第一作者的第一单位	上海科技大学
推荐引用方式 GB/T 7714	Huang, Shijie,Zhang, Kai,Huang, Jiawei,et al. Towards Accurate Fetal Brain Parcellation via Hierarchical Network and Loss[C]. GEWERBESTRASSE 11, CHAM, CH-6330, SWITZERLAND:Springer Science and Business Media Deutschland GmbH,2025:70-81.