Chinese Baby Connectome Project (CBCP): Advanced and Comprehensive Infant Cohort Protocols
Poster No:
1283
Submission Type:
Abstract Submission
Authors:
Lixuan Zhu1, Qing Yang1, Feiyu Quan1, Yan Liang1, Qiong He1, Ying Lin2, Jian Sun1, Ruili Ma1, Jiayang Guo1, Weijia Zhang1, Zhixin Wang1, Yajuan Zhang1, Tianli Tao1, Xinyi Cai1, Lin Zhang2, Mingwen Yang2, Zuozhen Lan2, Chaowei Zhang3, Dan Cai4, Na Hu4, Xianghui Huang5, Wei Wei6, Xuechen Ding4, Xinpei Xu6, Mingming Zhang4, Xiaoqian Yan7, Chenbo Wang8, Weijun Zhang9, Jiayu Zhu9, Jian Cheng10, Kunyu Xu11, Geng Chen12, Qian Wang1, Feng Shi13, Anqi Qiu14, Chunfeng Lian15, Jungang Liu2, Dan Li4, Guoqiang Cheng16, Han Zhang1,17, Dinggang Shen1,13,17
Institutions:
1School of Biomedical Engineering, ShanghaiTech University, Shanghai, China, 2Department of Radiology, Xiamen Children's Hospital, Children's Hospital, Fujian, China, 3Fujian Key Laboratory of Neonatal Diseases, Children's Hospital of Fudan University (Xiamen Branch), Fujian, China, 4School of Psychology, Shanghai Normal University, Shanghai, China, 5Fujian Key Laboratory of Neonatal Diseases, Children's Hospital of Fudan University (Xiamen Branch), Xiamen, Fujian, 6Shanghai Institute of Early Childhood Education, Shanghai Normal University, Shanghai, China, 7Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China, 8School of Psychology and Cognitive Science, East China Normal University, Shanghai, China, 9Shanghai United Imaging Healthcare Co., Ltd., Shanghai, China, 10School of Computer Science and Engineering, Beihang University, Beijing, China, 11Institute of Modern Languages and Linguistics, Fudan University, Shanghai, China, 12School of Computer Science and Engineering, Northwestern Polytechnical University, Xi'an, China, 13Shanghai United Imaging Intelligence Co., Ltd., Shanghai, China, 14Department of Biomedical Engineering, National University of Singapore, Singapore, 15School of Mathematics and Statistics, Xi'an Jiaotong University, Xi'an, China, 16Department of Neonatology, Children's Hospital of Fudan University, Fujian, China, 17Shanghai Clinical Research and Trail Center, Shanghai, China
First Author:
Co-Author(s):
Chaowei Zhang
Fujian Key Laboratory of Neonatal Diseases, Children's Hospital of Fudan University (Xiamen Branch)
Fujian, China
Fujian Key Laboratory of Neonatal Diseases, Children's Hospital of Fudan University (Xiamen Branch)
Fujian, China
Xianghui Huang
Fujian Key Laboratory of Neonatal Diseases, Children's Hospital of Fudan University (Xiamen Branch)
Xiamen, Fujian
Fujian Key Laboratory of Neonatal Diseases, Children's Hospital of Fudan University (Xiamen Branch)
Xiamen, Fujian
Xinpei Xu
Shanghai Institute of Early Childhood Education, Shanghai Normal University
Shanghai, China
Shanghai Institute of Early Childhood Education, Shanghai Normal University
Shanghai, China
Xiaoqian Yan
Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University
Shanghai, China
Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University
Shanghai, China
Chenbo Wang
School of Psychology and Cognitive Science, East China Normal University
Shanghai, China
School of Psychology and Cognitive Science, East China Normal University
Shanghai, China
Geng Chen
School of Computer Science and Engineering, Northwestern Polytechnical University
Xi'an, China
School of Computer Science and Engineering, Northwestern Polytechnical University
Xi'an, China
Han Zhang
School of Biomedical Engineering, ShanghaiTech University|Shanghai Clinical Research and Trail Center
Shanghai, China|Shanghai, China
School of Biomedical Engineering, ShanghaiTech University|Shanghai Clinical Research and Trail Center
Shanghai, China|Shanghai, China
Dinggang Shen
School of Biomedical Engineering, ShanghaiTech University|Shanghai United Imaging Intelligence Co., Ltd.|Shanghai Clinical Research and Trail Center
Shanghai, China|Shanghai, China|Shanghai, China
School of Biomedical Engineering, ShanghaiTech University|Shanghai United Imaging Intelligence Co., Ltd.|Shanghai Clinical Research and Trail Center
Shanghai, China|Shanghai, China|Shanghai, China
Introduction:
Studies based on the developing Human Connectome Project (dHCP) (Edwards et al., 2022) and the UNC/UMN Baby Connectome Project (BCP) (Howell et al., 2019) have significantly advanced our understanding of the early development of brain. Our research extends cutting-edge magnetic resonance imaging (MRI) acquisition techniques, including AI-driven approaches, to investigate infant brain development. In the Chinese Baby Connectome Project (CBCP), we integrate infant-tailored, accelerated imaging with comprehensive non-imaging data collection, aiming to build the first longitudinal atlas in China (Fig. 1b). Aligned with children aged 0-6 years, the CBCP protocol covers imaging, physiological, genetic, environmental, and behavioral data, providing a holistic perspective on brain development.
Methods:
Cohort Design: CBCP aims to gather full-spectrum data from over 1000 healthy infants aged 0-6 in 12 sub-cohorts, utilizing an accelerated longitudinal design. Rigorous criteria ensure enrollment of term-birth children without adverse prenatal conditions.
Preparations: Lab layout (Fig. 1a) and experimental procedure (Fig. 1c) are detailed. Pre-MRI preparation includes 25-min mock scanner training with motion monitoring, sleeping with noise training, and sleep deprivation to increase success rate.
MRI scan: we encourage children of age 0-3 undergo MRI scans during natural sleep and those of age 3-6 take awake scans while watching cartoons (black screen for resting-state fMRI (rfMRI)), but allow overlapped age period between sleep and awake scans. All sites use 3.0T MRI scanners of the same model (uMR 890, United Imaging) with a 64-channel head coil. Informed by pilot studies, CBCP MRI protocol (Table 1) encompass T1w and T2w structural MRI (sMRI), rfMRI, diffusion MRI (dMRI), and multiplex imaging (MTP) (Ye et al., 2022).
For sMRI, we utilize AI-assisted compressed sensing (ACS) (Liu et al., 2023) to ensure fast acquisition of 3D sMRI (0.8mm isotropic), saving 37.8% time compared to BCP. To further "freeze" head motion and increase success rate for certain infants who prone to move, 1-min ultra-fast 3D sMRI scans are implemented. Compared to BCP with 2mm isotropic rfMRI, CBCP collects rfMRI with higher spatial resolution (1.8mm isotropic) at a 800ms TR. dMRI with complete incremental acquisitions of 3-shell scheme at both AP and PA phase encoding directions. MTP incorporates dual repetition time, dual flip angle, multi-echo, and optional flow modulation features, which provides 16 qualitative and 9 quantitative images (Fig. 1b).
Quality control (QC): a pilot study with traveling subjects ensured data quality across centers. Phantom scans before the experiment assure scanner stability. All the MRI data will be immediately uploaded to a data-management platform for QC. An automated infant MRI preprocessing pipeline, using deep learning, guarantees reliable retrieval of subjects' attributes.
Non-imaging data: includes EEG/ERP for infants (0-6) during various tasks and resting state. Behavioral observations, audio/video recordings, social assessments, IQ, and Griffiths developmental evaluations are conducted following a comprehensive digitized questionnaire on environmental, perinatal, and family factors. Genetic data will be collected for deep sequencing.
Preparations: Lab layout (Fig. 1a) and experimental procedure (Fig. 1c) are detailed. Pre-MRI preparation includes 25-min mock scanner training with motion monitoring, sleeping with noise training, and sleep deprivation to increase success rate.
MRI scan: we encourage children of age 0-3 undergo MRI scans during natural sleep and those of age 3-6 take awake scans while watching cartoons (black screen for resting-state fMRI (rfMRI)), but allow overlapped age period between sleep and awake scans. All sites use 3.0T MRI scanners of the same model (uMR 890, United Imaging) with a 64-channel head coil. Informed by pilot studies, CBCP MRI protocol (Table 1) encompass T1w and T2w structural MRI (sMRI), rfMRI, diffusion MRI (dMRI), and multiplex imaging (MTP) (Ye et al., 2022).
For sMRI, we utilize AI-assisted compressed sensing (ACS) (Liu et al., 2023) to ensure fast acquisition of 3D sMRI (0.8mm isotropic), saving 37.8% time compared to BCP. To further "freeze" head motion and increase success rate for certain infants who prone to move, 1-min ultra-fast 3D sMRI scans are implemented. Compared to BCP with 2mm isotropic rfMRI, CBCP collects rfMRI with higher spatial resolution (1.8mm isotropic) at a 800ms TR. dMRI with complete incremental acquisitions of 3-shell scheme at both AP and PA phase encoding directions. MTP incorporates dual repetition time, dual flip angle, multi-echo, and optional flow modulation features, which provides 16 qualitative and 9 quantitative images (Fig. 1b).
Quality control (QC): a pilot study with traveling subjects ensured data quality across centers. Phantom scans before the experiment assure scanner stability. All the MRI data will be immediately uploaded to a data-management platform for QC. An automated infant MRI preprocessing pipeline, using deep learning, guarantees reliable retrieval of subjects' attributes.
Non-imaging data: includes EEG/ERP for infants (0-6) during various tasks and resting state. Behavioral observations, audio/video recordings, social assessments, IQ, and Griffiths developmental evaluations are conducted following a comprehensive digitized questionnaire on environmental, perinatal, and family factors. Genetic data will be collected for deep sequencing.
Results:
As of 10/23, we've collected MRI data from 298 infants with a 78% success rate (sMRI, fMRI, and dMRI complete acquisition). CBCP QC standards show pass rates: T1w, 94.4%; T2w, 85.7%; rfMRI, 69.4%; dMRI, 87.1%. Applying the same criteria to BCP data yields: T1w, 95.8%; T2w, 93.2%; rfMRI, 56%; dMRI, 72.1%.
Conclusions:
We report our protocol for this ongoing CBCP project aiming to build China's largest, high-quality, longitudinal infant brain-behavior dataset. CBCP embraces advanced imaging techniques for a comprehensive longitudinal cohort, offering a unified and standardized protocol from acquisition to analysis for both imaging and non-imaging data in depicting healthy brain developmental trajectories.
Lifespan Development:
Early life, Adolescence, Aging 2
Normal Brain Development: Fetus to Adolescence 1
Keywords:
Development
MRI
Other - Early Brain Development, Infant, AI-assisted Compressed Sensing，MULTIPLEX imaging
1|2Indicates the priority used for review
Provide references using author date format
Van Essen, D.C. et al. (2013) ‘The WU-Minn Human Connectome Project: An overview’, NeuroImage, 80, pp. 62–79. Available at: https://doi.org/10.1016/j.neuroimage.2013.05.041.
Howell, B.R. et al. (2019) ‘The UNC/UMN Baby Connectome Project (BCP): An overview of the study design and protocol development’, NeuroImage, 185, pp. 891–905. Available at: https://doi.org/10.1016/j.neuroimage.2018.03.049.
Liu, K. et al. (2023) ‘The clinical feasibility of artificial intelligence-assisted compressed sensing single-shot fluid-attenuated inversion recovery (ACS-SS-FLAIR) for evaluation of uncooperative patients with brain diseases: comparison with the conventional T2-FLAIR with parallel imaging’, Acta Radiologica, 64(5), pp. 1943–1949. Available at: https://doi.org/10.1177/02841851221139125.
Ye, Y. et al. (2022) ‘MULTI‐parametric MR imaging with fLEXible design (MULTIPLEX)’, Magnetic Resonance in Medicine, 87(2), pp. 658–673. Available at: https://doi.org/10.1002/mrm.28999.
Acknowledgement
This work is partially supported by the STI 2030—Major Projects (2022ZD0209000 and 2021ZD0200516), Shanghai Pilot Program for Basic Research—Chinese Academy of Science, Shanghai Branch (JCYJ-SHFY-2022-014), and Shenzhen Science and Technology Program (No. KCXFZ20211020163408012).