Abstract
In vivo fetal brain MRI is employed in clinical practice and in research studies to appreciate in utero brain development. There is increasing interest in transient regions of the fetal brain, such as the subplate (SP), ventricular zone (VZ) and ganglionic eminence (GE) (also referred to as germinal matrix), and their role in normal and abnormal antenatal brain development. On T1w and T2w fetal MRI, these transient regions are defined by highly heterogeneous and stratified signal intensities with rapidly changing patterns. In this work, we define the SP, VZ and GE in a 0.5mm isotropic resolution atlas from the developing Human Connectome Project (dHCP) [1, 17, 18], and train an Attention U-Net [12] to automatically segment them based on semi-automatically generated labels. Our solution spans from 21 through to 36 weeks gestational age (GA), offering insight into a crucial period of antenatal brain development. The proposed automated segmentation achieved mean Dice scores of 0.88, 0.70 and 0.82 for SP, VZ and GE respectively. A volumetric comparison of transient regions in a small cohort of fetuses with isolated ventriculomegaly (VM, n = 8) vs. controls (n = 265) showed significantly enlarged absolute volumes in the GE (P = 0.005) and VZ (P < 0.001) of the left hemisphere.
H.S. Sousa and A. Fukami-Gartner—Joint first authors H.S.S an A.F-G contributed equally to this work.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Developing human connectome project. http://www.developingconnectome.org
Mirtk software package. https://github.com/BioMedIA/MIRTK
Bayer, S.A., Altman, J.: The Human Brain During the Third Trimester. CRC Press (2003)
Bayer, S.A., Altman, J.: The human brain during the second trimester (2005)
Cardoso, M.J., et al.: Monai: An open-source framework for deep learning in healthcare. arXiv preprint arXiv:2211.02701 (2022)
Cordero-Grande, L., et al.: Automating motion compensation in 3T fetal brain imaging: localize, align and reconstruct. In: ISMRM 2019, p. 1000 (2019)
Gholipour, A., et al.: A normative spatiotemporal MRI atlas of the fetal brain for automatic segmentation and analysis of early brain growth. Sci. Rep. 7 (2017)
Griffiths, P.D., et al.: Sectional anatomy of the fetal brain (2010)
Jenkinson, M., et al.: FSl. Neuro Image 62(2), 782–790 (2012)
Kostović, I., et al.: Neural histology and neurogenesis of the human fetal and infant brain. Neuro Image 188, 743–773 (2019)
Machado-Rivas, F., et al.: Normal growth, sexual dimorphism, and lateral asymmetries at fetal brain MRI. Radiology 303, 162–170 (2022)
Oktay, O., et al.: Attention U-Net: learning where to look for the pancreas. In: MIDDL 2016 (2018)
Payette, K., et al.: An automatic multi-tissue human fetal brain segmentation benchmark using the fetal tissue annotation dataset. Sci. Data 8 (2021)
Price, A.N., et al.: The developing human connectome project (DHCP): fetal acquisition protocol. In: ISMRM 2019 (2019)
Rutherford, M., et al.: MR imaging methods for assessing fetal brain development (2008)
Stuempflen, M., et al.: The ganglionic eminence: volumetric assessment of transient brain structure utilizing fetal magnetic resonance imaging. Ultrasound in Obstet. & Gynecol. (2023)
Uus, A., et al.: Multi-channel spatio-temporal MRI atlas of the normal fetal brain development from the developing human connectome project. G-Node (2023)
Uus, A.U., et al.: Bounti: brain volumetry and automated parcellation for 3D fetal MRI. biorxiv (2023)
Vasung, L., et al.: Abnormal development of transient fetal zones in mild isolated fetal ventriculomegaly. Cereb. Cortex (2022)
Van der Walt, S., et al.: scikit-image: image processing in Python. PeerJ 2, e453 (2014)
Wright, R., et al.: LSTM spatial co-transformer networks for registration of 3D Fetal US and MR brain images. In: PIPPI/DATRA -2018. LNCS, vol. 11076, pp. 149–159. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-00807-9_15
Yushkevich, P.A., et al.: User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability. Neuro Image 31, 1116–1128 (2006)
Acknowledgments
We would like to thank all participating mothers & families and staff involved in the dHCP. AF-G.’s PhD research is funded by the MRC CNDD, and H.S.S by the EPSRC CDT in Smart Medical Imaging (EP/S022104/1). This research project was supported by the Academy of Medical Sciences Springboard Award (SBF0041040), ERC under the European Union’s Seventh Framework Programme [FP7/ 20072013]/ERC grant 319456 (dHCP), Wellcome/EPSRC Centre for Medical Engineering at KCL [WT 203148/Z/16/Z)], the NIHR CRF at Guy’s and St Thomas’. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Sousa, H.S. et al. (2023). A Deep Learning Approach for Segmenting the Subplate and Proliferative Zones in Fetal Brain MRI. In: Link-Sourani, D., Abaci Turk, E., Macgowan, C., Hutter, J., Melbourne, A., Licandro, R. (eds) Perinatal, Preterm and Paediatric Image Analysis. PIPPI 2023. Lecture Notes in Computer Science, vol 14246. Springer, Cham. https://doi.org/10.1007/978-3-031-45544-5_2
Download citation
DOI: https://doi.org/10.1007/978-3-031-45544-5_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-45543-8
Online ISBN: 978-3-031-45544-5
eBook Packages: Computer ScienceComputer Science (R0)