Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Protocol
  • Published:

iBEAT V2.0: a multisite-applicable, deep learning-based pipeline for infant cerebral cortical surface reconstruction


The human cerebral cortex undergoes dramatic and critical development during early postnatal stages. Benefiting from advances in neuroimaging, many infant brain magnetic resonance imaging (MRI) datasets have been collected from multiple imaging sites with different scanners and imaging protocols for the investigation of normal and abnormal early brain development. However, it is extremely challenging to precisely process and quantify infant brain development with these multisite imaging data because infant brain MRI scans exhibit (a) extremely low and dynamic tissue contrast caused by ongoing myelination and maturation and (b) inter-site data heterogeneity resulting from the use of diverse imaging protocols/scanners. Consequently, existing computational tools and pipelines typically perform poorly on infant MRI data. To address these challenges, we propose a robust, multisite-applicable, infant-tailored computational pipeline that leverages powerful deep learning techniques. The main functionality of the proposed pipeline includes preprocessing, brain skull stripping, tissue segmentation, topology correction, cortical surface reconstruction and measurement. Our pipeline can handle both T1w and T2w structural infant brain MR images well in a wide age range (from birth to 6 years of age) and is effective for different imaging protocols/scanners, despite being trained only on the data from the Baby Connectome Project. Extensive comparisons with existing methods on multisite, multimodal and multi-age datasets demonstrate superior effectiveness, accuracy and robustness of our pipeline. We have maintained a website, iBEAT Cloud, for users to process their images with our pipeline (, which has successfully processed over 16,000 infant MRI scans from more than 100 institutions with various imaging protocols/scanners.

This is a preview of subscription content, access via your institution

Access options

Buy this article

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Typical T1w and T2w infant brain MR images.
Fig. 2: Three 6-month-old brain images (different subjects) acquired with different scanners and imaging protocols.
Fig. 3: The framework of the iBEAT V2.0 computational pipeline.
Fig. 4: The T1w images, T2w images and reconstructed cortical surfaces for the same subject at different ages.
Fig. 5: Quantitative evaluation of GM and WM segmentation accuracy using different combinations of T1w and T2w images on MSMS6 dataset.
Fig. 6: Quantitative comparison of different methods on motion-free and motion-corrupted images.
Fig. 7: The evaluation of the pipeline motion artifacts robustness on typical BCP subjects of different ages.
Fig. 8: Quantitative comparison of segmentation between motion-free and motion-corrupted images with different ages and modalities.
Fig. 9: Quantitative comparison of segmentation performance with different image resolutions.
Fig. 10: The tissue contrast map and the quantitative comparison of segmentation performance in different cortical regions at different ages.
Fig. 11: The typical developmental trajectories measured on the BCP dataset.

Similar content being viewed by others

Data availability

To quantitatively validate our pipeline, we adopted three public datasets: the BCP dataset (, the dHCP dataset ( and a multisite multiscanner 6-month dataset, MSMS6 ( The software used in this protocol can be found at (iBEAT V2.0 Cloud) and iBEAT V2.0 Docker version (


  1. Gilmore, J. H., Knickmeyer, R. C. & Gao, W. Imaging structural and functional brain development in early childhood. Nat. Rev. Neurosci. 19, 123–137 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  2. Li, G. et al. Computational neuroanatomy of baby brains: a review. NeuroImage 185, 906–925 (2019).

    Article  PubMed  Google Scholar 

  3. Lyall, A. E. et al. Dynamic development of regional cortical thickness and surface area in early childhood. Cereb. Cortex 25, 2204–2212 (2015).

    Article  PubMed  Google Scholar 

  4. Garcia, K. E. et al. Dynamic patterns of cortical expansion during folding of the preterm human brain. Proc. Natl Acad. Sci. USA 115, 3156–3161 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Li, G. et al. Mapping longitudinal development of local cortical gyrification in infants from birth to 2 years of age. J. Neurosci. 34, 4228–4238 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Wang, F. et al. Developmental topography of cortical thickness during infancy. Proc. Natl Acad. Sci. USA 116, 15855–15860 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Shaw, P. et al. Intellectual ability and cortical development in children and adolescents. Nature 440, 676–679 (2006).

    Article  CAS  PubMed  Google Scholar 

  8. Howell, B. R. et al. The UNC/UMN Baby Connectome Project (BCP): an overview of the study design and protocol development. NeuroImage 185, 891–905 (2019).

    Article  PubMed  Google Scholar 

  9. Makropoulos, A. et al. The developing human connectome project: a minimal processing pipeline for neonatal cortical surface reconstruction. NeuroImage 173, 88–112 (2018).

    Article  PubMed  Google Scholar 

  10. Dale, A. M., Fischl, B. & Sereno, M. I. Cortical surface-based analysis: I. segmentation and surface reconstruction. NeuroImage 9, 179–194 (1999).

    Article  CAS  PubMed  Google Scholar 

  11. Fischl, B., Sereno, M. I. & Dale, A. M. Cortical surface-based analysis: II. inflation, flattening, and a surface-based coordinate system. NeuroImage 9, 195–207 (1999).

    Article  CAS  PubMed  Google Scholar 

  12. Hagler, D. J. et al. Image processing and analysis methods for the Adolescent Brain Cognitive Development Study. NeuroImage 202, 116091 (2019).

    Article  CAS  PubMed  Google Scholar 

  13. Glasser, M. F. et al. The minimal preprocessing pipelines for the Human Connectome Project. NeuroImage 80, 105–124 (2013).

    Article  PubMed  Google Scholar 

  14. Sun, Y. et al. Multi-site infant brain segmentation algorithms: the iSeg-2019 challenge. IEEE Trans. Med. Imaging 40, 1363–1376 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  15. Ellis, C. T. et al. Retinotopic organization of visual cortex in human infants. Neuron 109, 2616–2626.e6 (2021).

    Article  CAS  PubMed  Google Scholar 

  16. Hu, D. et al. Existence of functional connectome fingerprint during infancy and its stability over months. J. Neurosci. 42, 377–389 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Na, X. et al. Maternal obesity during pregnancy is associated with lower cortical thickness in the neonate brain. Am. J. Neuroradiol. (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  18. Wang, Y. et al. Developmental abnormalities of structural covariance networks of cortical thickness and surface area in autistic infants within the first 2 years. Cereb. Cortex (2022).

    Article  PubMed  PubMed Central  Google Scholar 

  19. Natu, V. S. et al. Infants’ cortex undergoes microstructural growth coupled with myelination during development. Commun. Biol. 4, 1–12 (2021).

    Article  Google Scholar 

  20. Grotheer, M. et al. White matter myelination during early infancy is linked to spatial gradients and myelin content at birth. Nat. Commun. 13, 1–12 (2022).

    Article  Google Scholar 

  21. Jiang, W. et al. Neural alterations in opioid-exposed infants revealed by edge-centric brain functional networks. Brain Commun. 4, (2022).

  22. Huang, Y. et al. Mapping developmental regionalization and patterns of cortical surface area from 29 post-menstrual weeks to 2 years of age. Proc. Natl Acad. Sci. USA 119, e2121748119 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Sled, J. G., Zijdenbos, A. P. & Evans, A. C. A nonparametric method for automatic correction of intensity nonuniformity in mri data. IEEE Trans. Med. Imaging 17, 87–97 (1998).

    Article  CAS  PubMed  Google Scholar 

  24. Tustison, N. J. et al. N4ITK: improved N3 bias correction. IEEE Trans. Med. Imaging 29, 1310–1320 (2010).

    Article  PubMed  PubMed Central  Google Scholar 

  25. Jenkinson, M. & Smith, S. A global optimisation method for robust affine registration of brain images. Med. Image Anal. 5, 143–156 (2001).

    Article  CAS  PubMed  Google Scholar 

  26. Jenkinson, M., Bannister, P., Brady, M. & Smith, S. Improved optimization for the robust and accurate linear registration and motion correction of brain images. NeuroImage 17, 825–841 (2002).

    Article  PubMed  Google Scholar 

  27. Sun, L. et al. Topological correction of infant white matter surfaces using anatomically constrained convolutional neural network. NeuroImage 198, 114–124 (2019).

    Article  PubMed  Google Scholar 

  28. Yeo, B. T. T. et al. Spherical demons: fast diffeomorphic landmark-free surface registration. IEEE Trans. Med. Imaging 29, 650–668 (2010).

    Article  PubMed  Google Scholar 

  29. Robinson, E. C. et al. MSM: a new flexible framework for multimodal surface matching. NeuroImage 100, 414–426 (2014).

    Article  PubMed  Google Scholar 

  30. Zhao, F. et al. Spherical deformable U-net: application to cortical surface parcellation and development prediction. IEEE Trans. Med. Imaging 40, 1217–1228 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  31. Zhao, F. et al. S3Reg: superfast spherical surface registration based on deep learning. IEEE Trans. Med. Imaging (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  32. Wang, L. et al. LINKS: Learning-based multi-source IntegratioN frameworK for Segmentation of infant brain images. NeuroImage 108, 160–172 (2015).

    Article  PubMed  Google Scholar 

  33. Zöllei, L., Iglesias, J. E., Ou, Y., Grant, P. E. & Fischl, B. Infant FreeSurfer: an automated segmentation and surface extraction pipeline for T1-weighted neuroimaging data of infants 0–2 years. NeuroImage 218, 116946 (2020).

    Article  PubMed  Google Scholar 

  34. Henschel, L. et al. FastSurfer—a fast and accurate deep learning based neuroimaging pipeline. NeuroImage 219, 117012 (2020).

    Article  PubMed  Google Scholar 

  35. Wang, L. et al. Benchmark on automatic six-month-old infant brain segmentation algorithms: the iSeg-2017 challenge. IEEE Trans. Med. Imaging 38, 2219–2230 (2019).

    Article  Google Scholar 

  36. Zaitsev, M., Maclaren, J. & Herbst, M. Motion artifacts in MRI: a complex problem with many partial solutions. J. Magn. Reson. Imaging 42, 887–901 (2015).

    Article  PubMed  Google Scholar 

  37. Paschal, C. B. & Morris, H. D. K-space in the clinic. J. Magn. Reson. Imaging 19, 145–159 (2004).

    Article  PubMed  Google Scholar 

  38. Drakulich, S. et al. Maturational trajectories of pericortical contrast in typical brain development. NeuroImage 235, (2021).

  39. Desikan, R. S. et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. NeuroImage 31, 968–980 (2006).

    Article  PubMed  Google Scholar 

  40. Wu, Z. et al. Construction of 4D infant cortical surface atlases with sharp folding patterns via spherical patch-based group-wise sparse representation. Hum. Brain Mapp. 40, 3860–3880 (2019).

    PubMed  PubMed Central  Google Scholar 

  41. Li, G. et al. Mapping region-specific longitudinal cortical surface expansion from birth to 2 years of age. Cereb. Cortex 23, 2724–2733 (2013).

    Article  PubMed  Google Scholar 

  42. Nie, J. et al. Longitudinal development of cortical thickness, folding, and fiber density networks in the first 2 years of life. Hum. Brain Mapp. 35, 3726–3737 (2014).

    Article  PubMed  Google Scholar 

  43. Li, G., Lin, W., Gilmore, J. H. & Shen, D. Spatial patterns, longitudinal development, and hemispheric asymmetries of cortical thickness in infants from birth to 2 years of age. J. Neurosci. 35, 9150–9162 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references


This work was partially supported by NIH grants (MH116225, MH117943, MH109773 and MH123202). This work also utilizes approaches developed by an NIH grant (1U01MH110274) and the efforts of the UNC/UMN Baby Connectome Project Consortium. The authors also thank D. Shen for an initial discussion of this work when he was with the University of North Carolina at Chapel Hill.

Author information

Authors and Affiliations



L.W. and G.L. designed the pipeline framework. L.W., Z.W. and G.L. implemented the pipeline and performed the validation. Z.W., L.C. and Y.S. prepared the experimental results. L.W., Z.W., W.L. and G.L. wrote the paper.

Corresponding authors

Correspondence to Li Wang, Zhengwang Wu or Gang Li.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Protocols thanks Jessica Dubois and Natasha Leporé for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Related links

Key references using this protocol

Ellis, C. T. et al. Neuron 109, 2616–2626.e6 (2021):

Natu, V. S. et al. Commun. Biol. 4, 1191 (2021):

Grotheer, M. et al. Nat. Commun. 13, 997 (2022):

Huang, Y. et al. Proc. Natl Acad. Sci. USA 119, e2121748119 (2022):

Extended data

Extended Data Fig. 1 Comparison of processing results between Infant FreeSurfer and iBEAT V2.0 on different datasets.

i. Comparison on the BCP dataset. ii. Comparison on the dHCP dataset. iii. Comparison on the MSMS dataset. (a) T1w images. (b) Tissue segmentation by Infant FreeSurfer. (c) Tissue segmentation by iBEAT V2.0. (d) iBEAT V2.0 reconstructed cortical surfaces overlayed on intensity images, with red contours indicating inner surfaces and green contours indicating outer surfaces. (e) Reconstructed inner cortical surface using Infant FreeSurfer. (f) Reconstructed inner cortical surfaces using iBEAT V2.0. (g) Reconstructed outer cortical surfaces (color-coded by cortical thickness) using iBEAT V2.0.

Extended Data Fig. 2 Comparison between dHCP released results and iBEAT V2.0 processed results.

i. Comparison between the dHCP released results and the iBEAT V2.0 processed results on two randomly selected subjects in the dHCP dataset. i.(a) T2w images; i.(b) Gray matter and white matter tissue boundaries from iBEAT V2.0 segmentation (green) and the dHCP released segmentation (red); i.(c) Color-coded vertex-wise surface distance maps between iBEAT V2.0 reconstructed inner surfaces and dHCP inner surfaces; i.(d) Color-coded vertex-wise surface distance maps between our reconstructed outer surfaces and dHCP outer surfaces; i.(e) Close-up views of our reconstructed inner and outer surfaces in i.(c) and i.(d); i.(f) Close-up views of the dHCP released inner and outer surfaces in i.(c) and i.(d). ii. Quantitative comparison between iBEAT V2.0 results and the dHCP released results. ii.(a) The Dice ratio of the gray matter and white matter; ii.(b) The average surface distance (ASD, mm) of the gray matter and white matter. iii. Comparison of the reconstructed cortical surfaces (colored-coded by mean curvature) in the occipital lobe of a typical subject using the dHCP pipeline and iBEAT V2.0 pipeline. iii.(a) The posterior view. iii.(b) The medial view of the left hemisphere. iii. (c) The medial view of the right hemisphere.

Extended Data Fig. 3 Comparison of the processed results using the FastSurfer pipeline and iBEAT V2.0 pipeline on different subjects (of different ages) from the BCP dataset.

(a). T1w images. (b) FastSurfer processed results. (c) iBEAT V2.0 processed results. The green contours in the figure indicate gray matter boundaries, and the red contours indicate white matter boundaries. Since FastSurfer is mainly developed for adult brains, two comparison pipelines achieved comparable results for the 60-month-old subject.

Extended Data Fig. 4 Quantitative evaluation of different methods on datasets from different scanners/protocols.

(a) The gray matter segmentation evaluation. (b) The white matter segmentation evaluation. The Dice ratio and Average Surface Distance (ASD, mm) evaluation metrics are reported for each evaluation. The error bars denote the standard deviation of the evaluation metric using different methods on the testing subjects.

Extended Data Fig. 5 Comparison of segmentation results by typical pipelines for 5 images from the BCP dataset with and without motion artifacts.

(a) Results for images without motion artifacts. (b) Results for images with motion artifacts. (c) Difference maps of the segmentation results with and without motion artifacts. Since IFS, volBrain, FreeSurfer, and FastSurfer mainly use T1w images as input, we also only used T1w images for a fair comparison.

Extended Data Fig. 6 Typical reconstructed cortical surfaces with different levels of quality.

Illustration of the different quality levels of reconstructed cortical surfaces. The first row shows the typical surfaces that are categorized as poor quality. These surfaces have large gyri or sulci missing, causing wrong geometry and/or wrong topology. The second row shows the surfaces with fair quality. They are generally accurately reconstructed in most brain regions, with tiny gyri/sulci missing. The third row shows the good surfaces with accurately reconstructed gyri and sulci.

Supplementary information

Supplementary Information

Supplementary methods, results and Figs. 1–8.

Reporting Summary

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wang, L., Wu, Z., Chen, L. et al. iBEAT V2.0: a multisite-applicable, deep learning-based pipeline for infant cerebral cortical surface reconstruction. Nat Protoc 18, 1488–1509 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing