Skip to main content
SpringerLink
Log in

Deep learning–based segmentation of whole-body fetal MRI and fetal weight estimation: assessing performance, repeatability, and reproducibility

  • Imaging Informatics and Artificial Intelligence
  • Published:
European Radiology Aims and scope Submit manuscript

Abstract

Objectives

To develop a deep-learning method for whole-body fetal segmentation based on MRI; to assess the method’s repeatability, reproducibility, and accuracy; to create an MRI-based normal fetal weight growth chart; and to assess the sensitivity to detect fetuses with growth restriction (FGR).

Methods

Retrospective data of 348 fetuses with gestational age (GA) of 19–39 weeks were included: 249 normal appropriate for GA (AGA), 19 FGR, and 80 Other (having various imaging abnormalities). A fetal whole-body segmentation model with a quality estimation module was developed and evaluated in 169 cases. The method was evaluated for its repeatability (repeated scans within the same scanner, n = 22), reproducibility (different scanners, n = 6), and accuracy (compared with birth weight, n = 7). A normal MRI-based growth chart was derived.

Results

The method achieved a Dice = 0.973, absolute volume difference ratio (VDR) = 1.8% and VDR mean difference = 0.75% (\({CI}_{95\%}\): − 3.95%, 5.46), and high agreement with the gold standard. The method achieved a repeatability coefficient = 4.01%, ICC = 0.99, high reproducibility with a mean difference = 2.21% (\({CI}_{95\%}\): − 1.92%, 6.35%), and high accuracy with a mean difference between estimated fetal weight (EFW) and birth weight of − 0.39% (\({CI}_{95\%}\): − 8.23%, 7.45%). A normal growth chart (n = 246) was consistent with four ultrasound charts. EFW based on MRI correctly predicted birth-weight percentiles for all 18 fetuses ≤ 10thpercentile and for 14 out of 17 FGR fetuses below the 3rd percentile. Six fetuses referred to MRI as AGA were found to be < 3rd percentile.

Conclusions

The proposed method for automatic MRI-based EFW demonstrated high performance and sensitivity to identify FGR fetuses.

Clinical relevance statement

Results from this study support the use of the automatic fetal weight estimation method based on MRI for the assessment of fetal development and to detect fetuses at risk for growth restriction.

Key Points

• An AI-based segmentation method with a quality assessment module for fetal weight estimation based on MRI was developed, achieving high repeatability, reproducibility, and accuracy.

• An MRI-based fetal weight growth chart constructed from a large cohort of normal and appropriate gestational-age fetuses is proposed.

• The method showed a high sensitivity for the diagnosis of small fetuses suspected of growth restriction.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Similar content being viewed by others

Abbreviations

AGA:

Appropriate for gestational age

aVDR :

Absolute volume difference ratio

EFW:

Estimated fetal weight

FGR:

Fetal growth restriction

FIESTA:

Fast imaging employing steady-state acquisition

GA:

Gestational age

HASTE:

Half-Fourier acquisition single-shot turbo spin-echo

ICC:

Intraclass correlation coefficient

TRUFI:

True fast imaging with steady-state free precession

TTA:

Test time augmentations

VDR :

Volume difference ratio

References

  1. Zhang-Rutledge K, Mack LM, Mastrobattista JM, Gandhi M (2018) Significance and outcomes of fetal growth restriction below the 5th percentile compared to the 5th to 10th percentiles on midgestation growth ultrasonography. J Ultrasound Med 37(9):2243–2249. https://doi.org/10.1002/jum.14577

    Article  PubMed  Google Scholar 

  2. Rizzo G, Mappa I, Bitsadze V et al (2020) Role of Doppler ultrasound at time of diagnosis of late-onset fetal growth restriction in predicting adverse perinatal outcome: prospective cohort study. Ultrasound Obstet Gynecol 55(6):793–798. https://doi.org/10.1002/uog.20406

    Article  CAS  PubMed  Google Scholar 

  3. Peleg D, Kennedy CM, Hunter SK (1998) Intrauterine growth restriction: identification and management. Am Fam Physician 58(2):453

    CAS  PubMed  Google Scholar 

  4. Milner J, Jane A (2018) The accuracy of ultrasound estimation of fetal weight in comparison to birth weight: a systematic review. Ultrasound 26(1):32–41. https://doi.org/10.1177/1742271X17732807

    Article  PubMed  PubMed Central  Google Scholar 

  5. Kacem Y, Cannie MM, Kadji C et al (2013) Fetal weight estimation: comparison of two-dimensional US and MR imaging assessments. Radiology 267(3):902–910. https://doi.org/10.1148/radiol.12121374

    Article  PubMed  Google Scholar 

  6. Sánchez-Fernández M, Corral ME, Aceituno L, Mazheika M, Mendoza N, Mozas-Moreno J (2021) Observer influence with other variables on the accuracy of Ultrasound estimation of fetal weight at term. Medicina (Kaunas) 57(3):216. https://doi.org/10.3390/medicina57030216

  7. Davidson JR, Uus A, Matthew J et al (2021) Fetal body MRI and its application to fetal and neonatal treatment: an illustrative review. Lancet Child Adolesc Health 5(6):447–458. https://doi.org/10.1016/S2352-4642(20)30313-8

    Article  PubMed  PubMed Central  Google Scholar 

  8. Gholipour A, Estroff JA, Barnewolt CE et al (2014) Fetal MRI: a technical update with educational aspirations. Concepts Magn Reson Part A Bridg Educ Res 43(6):237–266. https://doi.org/10.1002/cmr.a.21321

  9. Uotila J, Dastidar P, Heinonen T, Ryymin P, Punnonen R, Laasonen E (2000) Magnetic resonance imaging compared to ultrasonography in fetal weight and volume estimation in diabetic and normal pregnancy. Acta Obstet Gynecol Scand 79(4):255–259. https://doi.org/10.1034/j.1600-0412.2000.079004255.x

    Article  CAS  PubMed  Google Scholar 

  10. Zaretsky MV, Reichel TF, McIntire DD, Twickler DM (2003) Comparison of magnetic resonance imaging to ultrasound in the estimation of birth weight at term. Am J Obstet Gynecol 189(4):1017–1020. https://doi.org/10.1067/S0002-9378(03)00895-0

    Article  PubMed  Google Scholar 

  11. Kadji C, Cannie MM, Resta S et al (2019) Magnetic resonance imaging for prenatal estimation of birthweight in pregnancy: review of available data, techniques, and future perspectives. Am J Obstet Gynecol 220(5):428–439. https://doi.org/10.1016/j.ajog.2018.12.031

    Article  PubMed  Google Scholar 

  12. Torrents-Barrena J, Piella G, Masoller N et al (2019) Segmentation and classification in MRI and US fetal imaging: recent trends and future prospects. Med Image Anal 1(51):61–88. https://doi.org/10.1016/j.media.2018.10.003

    Article  Google Scholar 

  13. Anquez J, Bibin L, Angelini ED, Bloch I (2010) Segmentation of the fetal envelope on ante-natal MRI. Proc IEEE Int Symposium on Biomedical Imaging, pp 896–899. https://doi.org/10.1109/ISBI.2010.5490131

  14. Zhang T, Matthew J, Lohezic M et al (2016) Graph-based whole body segmentation in fetal MR images. Proc MICCAI Work PIPPI, Athens, Greece 21:21

    Google Scholar 

  15. Dudovitch G, Link-Sourani D, Ben Sira L, Miller E, Ben Bashat D, Joskowicz L (2020) Deep learning automatic fetal structures segmentation in MRI scans with few annotated datasets. In Proc. International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, Cham, pp. 365–374. https://doi.org/10.1007/978-3-030-59725-2_35

  16. Lo J, Nithiyanantham S, Cardinell J et al (2021) Cross Attention Squeeze Excitation Network (CASE-Net) for whole body fetal MRI segmentation. Sensors (Basel) 21(13):4490. https://doi.org/10.3390/s21134490

  17. Ryd D, Nilsson A, Heiberg E, Hedstrom E (2022) Automatic segmentation of the fetus in 3D magnetic resonance images using deep learning: accurate and fast fetal volume quantification for clinical use. Pediatr Cardiol. https://doi.org/10.1007/s00246-022-03038-0

    Article  PubMed  PubMed Central  Google Scholar 

  18. Bartlett JW, Frost C (2008) Reliability, repeatability and reproducibility: analysis of measurement errors in continuous variables. Ultrasound Obstet Gynecol 31(4):466–475. https://doi.org/10.1002/uog.5256

  19. Specktor-Fadida B, Link-Sourani D, Ferster-Kveller S et al (2021) A bootstrap self-training method for sequence transfer: state-of-the-art placenta segmentation in fetal MRI. In Proc. Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, and Perinatal Imaging, Placental and Preterm Image Analysis. Springer, Cham, pp 189–199. https://doi.org/10.1007/978-3-030-87735-4_18

  20. Isensee F, Jaeger PF, Kohl SA, Petersen J, Maier-Hein KH (2021) nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation. Nat Methods 18(2):203–211. https://doi.org/10.1038/s41592-020-01008-z

    Article  CAS  PubMed  Google Scholar 

  21. Wang G, Li W, Aertsen M, Deprest J, Ourselin S, Vercauteren T (2019) Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networks. Neurocomputing 21(338):34–45. https://doi.org/10.1016/j.neucom.2019.01.103

    Article  Google Scholar 

  22. Daniel-Spiegel E, Weiner E, Yarom I et al (2013) Establishment of fetal biometric charts using quantile regression analysis. J Ultrasound Med 32(1):23–33. https://doi.org/10.7863/jum.2013.32.1.23

    Article  PubMed  Google Scholar 

  23. Hadlock FP, Harrist RB, Martinez-Poyer J (1991) In utero analysis of fetal growth: a sonographic weight standard. Radiology 181(1):129–133. https://doi.org/10.1148/radiology.181.1.1887021

    Article  CAS  PubMed  Google Scholar 

  24. Kiserud T, Piaggio G, Carroli G et al (2017) The World Health Organization fetal growth charts: a multinational longitudinal study of ultrasound biometric measurements and estimated fetal weight. PLoS Med 14(1):e1002220. https://doi.org/10.1371/journal.pmed.1002220

    Article  PubMed  PubMed Central  Google Scholar 

  25. Stirnemann J, Villar J, Salomon LJ et al (2017) International estimated fetal weight standards of the INTERGROWTH-21st Project. Ultrasound Obstet Gynecol 49(4):478–486. https://doi.org/10.1002/uog.17347

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Nicolaides KH, Wright D, Syngelaki A, Wright A, Akolekar R (2018) Fetal Medicine Foundation fetal and neonatal population weight charts. Ultrasound Obstet Gynecol 52(1):44–51. https://doi.org/10.1002/uog.19073

    Article  CAS  PubMed  Google Scholar 

  27. Dudley NJ (2005) A systematic review of the ultrasound estimation of fetal weight. Ultrasound Obstet Gynecol 25(1):80–89. https://doi.org/10.1002/uog.1751

  28. Raunig DL, McShane LM, Pennello G et al (2015) Quantitative imaging biomarkers: a review of statistical methods for technical performance assessment. Stat Methods Med Res 24(1):27–67. https://doi.org/10.1177/0962280214537344

    Article  MathSciNet  PubMed  Google Scholar 

  29. Joskowicz L, Cohen D, Caplan N, Sosna J (2019) Inter-observer variability of manual contour delineation of structures in CT. Eur Radiol 29(3):1391–1399. https://doi.org/10.1007/s00330-018-5695-5

    Article  PubMed  Google Scholar 

  30. Liao K, Tang L, Peng C et al (2019) A modified model can improve the accuracy of fetal weight estimation by magnetic resonance imaging. Eur J Radiol 1(110):242–248. https://doi.org/10.1016/j.ejrad.2018.12.009

    Article  Google Scholar 

  31. Schild RL (2007) Three-dimensional volumetry and fetal weight measurement. Ultrasound Obstet Gynecol 30(6):799–803. https://doi.org/10.1002/uog.5181

    Article  CAS  PubMed  Google Scholar 

  32. Baker PN, Johnson IR, Gowland PA et al (1994) Fetal weight estimation by echo-planar magnetic resonance imaging. Lancet 343(8898):644–645. https://doi.org/10.1016/S0140-6736(94)92638-7

    Article  CAS  PubMed  Google Scholar 

  33. Shinozuka N, Okai T, Kohzuma S et al (1987) Formulas for fetal weight estimation by ultrasound measurements based on neonatal specific gravities and volumes. Am J Obstet Gynecol 157(5):1140–1145. https://doi.org/10.1016/S0002-9378(87)80278-8

    Article  CAS  PubMed  Google Scholar 

  34. McCrindle B, Zukotynski K, Doyle TE, Noseworthy MD (2021) A radiology-focused review of predictive uncertainty for AI interpretability in computer-assisted segmentation. Radiol Artif Intell 3(6). https://doi.org/10.1148/ryai.2021210031

Download references

Acknowledgements

This research was supported by Kamin grants 63418 and 72126 from the Israel Innovation Authority. We are grateful to the MRI technicians for scanning the fetuses, to the mothers who participated, and to those involved in data preparation and annotation: Ori Benzvi helped with data extraction, and Tuvia Ganot, Kerina Krupnik, Cassandra Kapoor, and Shelley Levi helped with the annotation process and data transfer.

Funding

This work was financially supported by Kamin grants of the Israel Innovation Authority.

Author information

Authors and Affiliations

  1. School of Computer Science and Engineering, The Hebrew University of Jerusalem, Jerusalem, Israel

    Bella Specktor-Fadida, Netanell Avisdris & Leo Joskowicz

  2. Sagol Brain Institute, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

    Daphna Link-Sourani, Aviad Rabinowich, Anna Levchakov, Netanell Avisdris & Dafna Ben-Bashat

  3. Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel

    Aviad Rabinowich, Liat Ben-Sira, Liran Hiersch & Dafna Ben-Bashat

  4. Department of Radiology, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

    Aviad Rabinowich & Liat Ben-Sira

  5. Department of Medical Imaging, The Hospital for Sick Children, University of Toronto, Toronto, Canada

    Elka Miller

  6. Department of Medical Imaging, CHEO, University of Ottawa, Ottawa, Canada

    Elka Miller

  7. Sagol School of Neuroscience, Tel Aviv University, Tel Aviv, Israel

    Liat Ben-Sira & Dafna Ben-Bashat

  8. Department of Obstetrics and Gynecology, Lis Hospital for Women’s Health, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

    Liran Hiersch

  9. Edmond and Lily Safra Center for Brain Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel

    Leo Joskowicz

Authors
  1. Bella Specktor-Fadida
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Aviad Rabinowich
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Elka Miller
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Anna Levchakov
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Netanell Avisdris
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Liat Ben-Sira
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Liran Hiersch
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Leo Joskowicz
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Dafna Ben-Bashat
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding authors

Correspondence to Bella Specktor-Fadida or Dafna Ben-Bashat.

Ethics declarations

Guarantor

The scientific guarantor of this publication is Dafna Ben-Bashat.

Conflict of interest

The authors of this manuscript declare no relationships with any companies whose products or services may be related to the subject matter of the article.

Statistics and biometry

One of the authors has significant statistical expertise.

Informed consent

Written informed consent was waived by the Institutional Review Boards of Tel Aviv Sourasky Medical Center and Children’s Hospital of Eastern Ontario for the retrospective data and was obtained from all participants in the prospective data at Tel Aviv Sourasky Medical Center.

Ethical approval

Institutional Review Board approval was obtained.

Study subjects or cohorts overlap

Study subjects have not been previously reported.

Methodology

• prospective and retrospective

• cross-sectional study

• multicenter study

Additional information

Publisher's note

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

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 420 KB)

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

Specktor-Fadida, B., Link-Sourani, D., Rabinowich, A. et al. Deep learning–based segmentation of whole-body fetal MRI and fetal weight estimation: assessing performance, repeatability, and reproducibility. Eur Radiol 34, 2072–2083 (2024). https://doi.org/10.1007/s00330-023-10038-y

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00330-023-10038-y

Keywords

Search

Navigation