European Radiology

, Volume 29, Issue 12, pp 6805–6815 | Cite as

Estimation of the radiation dose in pregnancy: an automated patient-specific model using convolutional neural networks

  • Tianwu Xie
  • Habib ZaidiEmail author



The conceptus dose during diagnostic imaging procedures for pregnant patients raises health concerns owing to the high radiosensitivity of the developing embryo/fetus. The aim of this work is to develop a methodology for automated construction of patient-specific computational phantoms based on actual patient CT images to enable accurate estimation of conceptus dose.


We developed a 3D deep convolutional network algorithm for automated segmentation of CT images to build realistic computational phantoms. The neural network architecture consists of analysis and synthesis paths with four resolution levels each, trained on manually labeled CT scans of six identified anatomical structures. Thirty-two CT exams were augmented to 128 datasets and randomly split into 80%/20% for training/testing. The absorbed doses for six segmented organs/tissues from abdominal CT scans were estimated using Monte Carlo calculations. The resulting radiation doses were then compared between the computational models generated using automated segmentation and manual segmentation, serving as reference.


The Dice similarity coefficient for identified internal organs between manual segmentation and automated segmentation results varies from 0.92 to 0.98 while the mean Hausdorff distance for the uterus is 16.1 mm. The mean absorbed dose for the uterus is 2.9 mGy whereas the mean organ dose differences between manual and automated segmentation techniques are 0.07%, − 0.45%, − 1.55%, − 0.48%, − 0.12%, and 0.28% for the kidney, liver, lung, skeleton, uterus, and total body, respectively.


The proposed methodology allows automated construction of realistic computational models that can be exploited to estimate patient-specific organ radiation doses from radiological imaging procedures.

Key Points

• The conceptus dose during diagnostic radiology and nuclear medicine imaging procedures for pregnant patients raises health concerns owing to the high radiosensitivity of the developing embryo/fetus.

• The proposed methodology allows automated construction of realistic computational models that can be exploited to estimate patient-specific organ radiation doses from radiological imaging procedures.

• The dosimetric results can be used for the risk-benefit analysis of radiation hazards to conceptus from diagnostic imaging procedures, thus guiding the decision-making process.


Multidetector-row computed tomography Radiologic phantoms Patient-specific computational modeling Radiation dosimetry 



Adaptive moment estimation


Convolutional neural networks


Computed tomography


Dice similarity coefficient


Hausdorff distance


Geneva University Hospital


Intelligence quotient


Positron emission tomography


Positron emission tomography/computed tomography


Positive predictive value


Rectified linear unit



This work was supported by the Swiss National Science Foundation under grant SNSF 320030_176052 and Qatar National Research Fund under grant NPRP10-0126-170263. No other potential conflicts of interest relevant to this article exist.


This study has received funding by the Swiss National Science Foundation.

Compliance with ethical standards


The scientific guarantor of this publication is Habib Zaidi.

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

No complex statistical methods were necessary for this paper.

Informed consent

Written informed consent was waived in this study.

Ethical approval

Institutional Review Board approval was obtained.


• Prospective

• Experimental

• Performed at one institution


  1. 1.
    Lazarus E, Debenedectis C, North D, Spencer PK, Mayo-Smith WW (2009) Utilization of imaging in pregnant patients: 10-year review of 5270 examinations in 3285 patients—1997–2006 Radiology 251:517–524CrossRefGoogle Scholar
  2. 2.
    Xie T, Zaidi H (2014) Fetal and maternal absorbed dose estimates for positron-emitting molecular imaging probes. J Nucl Med 55:1459–1166CrossRefGoogle Scholar
  3. 3.
    Zanotti-Fregonara P, Laforest R, Wallis JW (2015) Fetal radiation dose from [18F]-FDG in pregnant patients imaged with PET, PET/CT and PET/MRI. J Nucl Med 56:1218–1222CrossRefGoogle Scholar
  4. 4.
    Poletti PA, Botsikas D, Becker M et al (2019) Suspicion of appendicitis in pregnant women: emergency evaluation by sonography and low-dose CT with oral contrast. Eur Radiol 29:345–352CrossRefGoogle Scholar
  5. 5.
    American College of Radiology (2013) ACR-SPR practice parameter for imaging pregnant or potentially pregnant adolescents and women with ionizing radiation. Revised 2013 (Resolution 48)Google Scholar
  6. 6.
    ICRP (2000) ICRP publication 84: pregnancy and medical radiation. Ann ICRP 30:iii–viii, 1-43CrossRefGoogle Scholar
  7. 7.
    ICRP (2003) ICRP publication 90: biological effects after prenatal irradiation (embryo and fetus). Ann ICRP 33:5–206Google Scholar
  8. 8.
    Altman KI, Lett JT (2013) Relative radiation sensitivities of human organ systems. ElsevierGoogle Scholar
  9. 9.
    McCollough CH, Schueler BA, Atwell TD et al (2007) Radiation exposure and pregnancy: when should we be concerned? Radiographics 27:909–917 discussion 917-908CrossRefGoogle Scholar
  10. 10.
    Pearce MS, Salotti JA, Little MP et al (2012) Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet 380:499–505CrossRefGoogle Scholar
  11. 11.
    (2007) ICRP publication 105. Radiological protection in medicine. Ann ICRP 37:1–63Google Scholar
  12. 12.
    Vañó E, Miller DL, Martin CJ et al(2017) ICRP publication 135: diagnostic reference levels in medical imaging. Ann ICRP 46:1–144Google Scholar
  13. 13.
    Angel E, Wellnitz CV, Goodsitt MM et al (2008) Radiation dose to the fetus for pregnant patients undergoing multidetector CT imaging: Monte Carlo simulations estimating fetal dose for a range of gestational age and patient size. Radiology 249:220–227CrossRefGoogle Scholar
  14. 14.
    Damilakis J, Perisinakis K, Tzedakis A, Papadakis AE, Karantanas A (2010) Radiation dose to the conceptus from multidetector CT during early gestation: a method that allows for variations in maternal body size and conceptus position. Radiology 257:483–489CrossRefGoogle Scholar
  15. 15.
    Damilakis J, Tzedakis A, Perisinakis K, Papadakis AE (2010) A method of estimating conceptus doses resulting from multidetector CT examinations during all stages of gestation. Med Phys 37:6411–6420CrossRefGoogle Scholar
  16. 16.
    Helmrot E, Pettersson H, Sandborg M, Altén JN (2007) Estimation of dose to the unborn child at diagnostic X-ray examinations based on data registered in RIS/PACS. Eur Radiol 17:205–209CrossRefGoogle Scholar
  17. 17.
    Gu J, Xu XG, Caracappa PF, Liu B (2013) Fetal doses to pregnant patients from CT with tube current modulation calculated using Monte Carlo simulations and realistic phantoms. Radiat Prot Dosimetry 155:64–72CrossRefGoogle Scholar
  18. 18.
    Winer-Muram HT, Boone JM, Brown HL, Jennings SG, Mabie WC, Lombardo GT (2002) Pulmonary embolism in pregnant patients: fetal radiation dose with helical CT. Radiology 224:487–492CrossRefGoogle Scholar
  19. 19.
    Hoseinian-Azghadi E, Rafat-Motavalli L, Miri-Hakimabad H (2014) Development of a 9-months pregnant hybrid phantom and its internal dosimetry for thyroid agents. J Radiat Res 55:730–747CrossRefGoogle Scholar
  20. 20.
    Lopez-Rendon X, Walgraeve MS, Woussen S et al (2017) Comparing different methods for estimating radiation dose to the conceptus. Eur Radiol 27:851–858CrossRefGoogle Scholar
  21. 21.
    Maynard MR, Long NS, Moawad NS et al (2014) The UF family of hybrid phantoms of the pregnant female for computational radiation dosimetry. Phys Med Biol 59:4325–4343CrossRefGoogle Scholar
  22. 22.
    Xie T, Zaidi H (2016) Development of computational pregnant female and fetus models and assessment of radiation dose from positron-emitting tracers. Eur J Nucl Med Mol Imaging 43:2290–2300CrossRefGoogle Scholar
  23. 23.
    Xie T, Zanotti-Fregonara P, Edet-Sanson A, Zaidi H (2018) Patient-specific computational model and dosimetry calculations for a patient pregnant with twins undergoing a PET/CT examination. J Nucl Med 59:1451–1458CrossRefGoogle Scholar
  24. 24.
    Gilet AG, Dunkin JM, Fernandez TJ, Button TM, Budorick NE (2011) Fetal radiation dose during gestation estimated on an anthropomorphic phantom for three generations of CT scanners. AJR Am J Roentgenol 196:1133–1137CrossRefGoogle Scholar
  25. 25.
    Chatterson LC, Leswick DA, Fladeland DA, Hunt MM, Webster S, Lim H (2014) Fetal shielding combined with state of the art CT dose reduction strategies during maternal chest CT. Eur J Radiol 83:1199–1204CrossRefGoogle Scholar
  26. 26.
    Jaffe TA, Neville AM, Anderson-Evans C et al (2009) Early first trimester fetal dose estimation method in a multivendor study of 16- and 64-MDCT scanners and low-dose imaging protocols. AJR Am J Roentgenol 193:1019–1024CrossRefGoogle Scholar
  27. 27.
    Solomou G, Papadakis AE, Damilakis J (2015) Abdominal CT during pregnancy: a phantom study on the effect of patient centering on conceptus radiation dose and image quality. Eur Radiol 25:911–921CrossRefGoogle Scholar
  28. 28.
    Kelaranta A, Kaasalainen T, Seuri R, Toroi P, Kortesniemi M (2015) Fetal radiation dose in computed tomography. Radiat Prot Dosimetry 165:226–230CrossRefGoogle Scholar
  29. 29.
    Zaidi H, Xu XG (2007) Computational anthropomorphic models of the human anatomy: the path to realistic Monte Carlo modeling in medical imaging. Annu Rev Biomed Eng 9:471–500CrossRefGoogle Scholar
  30. 30.
    Zaidi H, Ay MR (2007) Current status and new horizons in Monte Carlo simulation of X-ray CT scanners. Med Biol Eng Comput 45:809–817CrossRefGoogle Scholar
  31. 31.
    Xie T, Poletti PA, Platon A, Becker CD, Zaidi H (2018) Assessment of CT dose to the fetus and pregnant female patient using patient-specific computational models. Eur Radiol 28:1054–1065CrossRefGoogle Scholar
  32. 32.
    Litjens G, Kooi T, Bejnordi BE et al (2017) A survey on deep learning in medical image analysis. Med Image Anal 42:60–88CrossRefGoogle Scholar
  33. 33.
    Mikolov T, Karafiát M, Burget L, Černocký J, Khudanpur S (2010) Recurrent neural network based language model. 11th annual conference of the International Speech Communication Association (INTERSPEECH-2010), Makuhari, Chiba, Japan, September 26–30. 2010, pp 1045–1048Google Scholar
  34. 34.
    Chung J, Gulcehre C, Cho K, Bengio Y (2015) Gated feedback recurrent neural networks. Proceedings of the 32nd International Conference on International Conference on Machine Learning. J PMLR 37:2067–2075Google Scholar
  35. 35.
    Mohamed A, Dahl GE, Hinton G (2012) Acoustic modeling using deep belief networks. IEEE Trans Audio Speech Lang Process 20:14–22CrossRefGoogle Scholar
  36. 36.
    Nasrabadi NM, Choo CY (1992) Hopfield network for stereo vision correspondence. IEEE Trans Neural Netw 3:5–13CrossRefGoogle Scholar
  37. 37.
    Emami H, Dong M, Nejad-Davarani SP, Glide-Hurst CK (2018) Generating synthetic CTs from magnetic resonance images using generative adversarial networks. Med Phys 45:3627–3636CrossRefGoogle Scholar
  38. 38.
    Sum J, Leung CS (2019) Learning algorithm for Boltzmann machines with additive weight and bias noise. IEEE Trans Neural Netw Learn Syst.
  39. 39.
    Krizhevsky A, Sutskever I, Hinton GE (2017) ImageNet classification with deep convolutional neural networks. Commun ACM 60:84–90CrossRefGoogle Scholar
  40. 40.
    Zhou X, Takayama R, Wang S, Zhou X, Hara T, Fujita H (2017) Automated segmentation of 3D anatomical structures on CT images by using a deep convolutional network based on end-to-end learning approach. Medical imaging: image processingGoogle Scholar
  41. 41.
    Liu X, Guo S, Yang B et al (2018) Automatic organ segmentation for CT scans based on super-pixel and convolutional neural networks. J Digit Imaging 31:748–760CrossRefGoogle Scholar
  42. 42.
    Weston AD, Korfiatis P, Kline TL et al (2018) Automated abdominal segmentation of CT scans for body composition analysis using deep learning. Radiology 290:669–679CrossRefGoogle Scholar
  43. 43.
    Ronneberger O, Fischer P, Brox T (2015) U-Net: convolutional networks for biomedical image segmentation. In: Navab N, Hornegger J, Wells W, Frangi A (eds) Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. MICCAI 2015. Lecture Notes in Computer Science, vol 9351. Springer, Cham Springer, pp 234–241Google Scholar
  44. 44.
    Çiçek Ö, Abdulkadir A, Lienkamp SS, Brox T, Ronneberger O (2016) 3D U-Net: learning dense volumetric segmentation from sparse annotation. In: Ourselin S, Joskowicz L, Sabuncu M, Unal G, Wells W (eds) Medical Image Computing and Computer-Assisted Intervention – MICCAI 2016. MICCAI 2016. Lecture Notes in Computer Science, vol 9901. Springer, Cham, pp 424–432Google Scholar
  45. 45.
    Chollet F (2018) Keras: the Python deep learning library. Astrophysics Source Code Library. Available at
  46. 46.
    (2002) Basic anatomical and physiological data for use in radiological protection: reference values. A report of age- and gender-related differences in the anatomical and physiological characteristics of reference individuals. ICRP Publication 89. Ann ICRP 32:5–265Google Scholar
  47. 47.
    Akbarzadeh A, Ay MR, Ghadiri H, Sarkar S, Zaidi H (2010) Measurement of scattered radiation in a volumetric 64-slice CT scanner using three experimental techniques. Phys Med Biol 55:2269–2280CrossRefGoogle Scholar
  48. 48.
    Pelowitz DB (2005) MCNPX user’s manual Version 2.5.0. Los Alamos National Laboratory, Los AlamosGoogle Scholar
  49. 49.
    Stovall M, Blackwell CR, Cundiff J et al (1995) Fetal dose from radiotherapy with photon beams: report of AAPM Radiation Therapy Committee Task Group No. 36. Med Phys 22:63–82CrossRefGoogle Scholar

Copyright information

© European Society of Radiology 2019

Authors and Affiliations

  1. 1.Division of Nuclear Medicine and Molecular ImagingGeneva University HospitalGeneva 4Switzerland
  2. 2.Geneva University NeurocenterGeneva UniversityGenevaSwitzerland
  3. 3.Department of Nuclear Medicine and Molecular ImagingUniversity of Groningen, University Medical Center GroningenGroningenNetherlands
  4. 4.Department of Nuclear MedicineUniversity of Southern DenmarkOdenseDenmark

Personalised recommendations