Classification of Foetal Distress and Hypoxia Using Machine Learning Approaches

  • Rounaq Abbas
  • Abir Jaafar HussainEmail author
  • Dhiya Al-Jumeily
  • Thar Baker
  • Asad Khattak
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10956)


Foetal distress and hypoxia (oxygen deprivation) is considered as a serious condition and one of the main factors for caesarean section in the obstetrics and Gynecology department. It is the third most common cause of death in new-born babies. Many foetuses that experienced some sort of hypoxic effects can develop series risks including damage to the cells of the central nervous system that may lead to life-long disability (cerebral palsy) or even death. Continuous labour monitoring is essential to observe the foetal well being. Foetal surveillance by monitoring the foetal heart rate with a cardiotocography is widely used. Despite the indication of normal results, these results are not reassuring, and a small proportion of these foetuses are actually hypoxic. In this paper, machine-learning algorithms are utilized to classify foetuses which are experiencing oxygen deprivation using PH value (a measure of hydrogen ion concentration of blood used to specify the acidity or alkalinity) and Base Deficit of extra cellular fluid level (a measure of the total concentration of blood buffer base that indicates the metabolic acidosis or compensated respiratory alkalosis) as indicators of respiratory and metabolic acidosis, respectively, using open source partum clinical data obtained from Physionet. Six well know machine learning classifier models are utilised in our experiments for the evaluation; each model was presented with a set of selected features derived from the clinical data. Classifier’s evaluation is performed using the receiver operating characteristic curve analysis, area under the curve plots, as well as the confusion matrix. Our simulation results indicate that machine-learning algorithms provide viable methods that could delivery improvements over conventional analysis.


Machine learning Hypoxia Foetal distress 



The authors would like to thanks Liverpool John Moores University for the scholarship to complete this research. In addition, this research work was partially supported by Zayed University Research Cluster Award # R18038.


  1. 1.
    Talaulikar, V.S., Arulkumaran, S.: Maternal, perinatal and long-term outcomes after assisted reproductive techniques (ART): implications for clinical practice. Eur. J. Obstet. Gynaecol. Reprod. Biol. 170(1), 13–19 (2013)CrossRefGoogle Scholar
  2. 2.
    Bobrow, C.S., Soothill, P.W.: Causes and consequences of fetal acidosis. Arch. Dis. Child. Fetal Neonatal Edition 80(3), F246–F249 (1999)CrossRefGoogle Scholar
  3. 3.
    Hasan, M.A., Reaz, M.B.I., Ibrahimy, M.I., Hussain, M.S., Uddin, J.: Detection and processing techniques of FECG signal for fetal monitoring. Biol. Proced. Online 11(1), 263 (2009)CrossRefGoogle Scholar
  4. 4.
    Talaulikar, V.S., Arulkumaran, S.: Persistent challenge of intrapartum fetal heart rate monitoring. Dasgupta’s Recent Adv. Obstet. Gynecol. 9, 68 (2012)Google Scholar
  5. 5.
    Wiberg-Itzel, E., Lipponer, C., Norman, M., Herbst, A., Prebensen, D., Hansson, A., Bryngelsson, A.L., Christoffersson, M., Sennström, M., Wennerholm, U.B., Nordström, L.: Determination of pH or lactate in fetal scalp blood in management of intrapartum fetal distress: randomised controlled multicentre trial. BMJ 336(7656), 1284–1287 (2008)CrossRefGoogle Scholar
  6. 6.
    Vintzileos, A.M., Nochimson, D.J., Antsaklis, A., Varvarigos, I., Guzman, I., Knuppel, R.A.: Comparison of intrapartum electronic fetal heart monitoring versus intermittent auscultation in detecting fetal acidemia at birth. Am. J. Obstet. Gynecol. 173, 1021–1024 (1995)CrossRefGoogle Scholar
  7. 7.
    Ingemarsson, I., Ingemarsson, E., Spencer, J.A.D.: Fetal Heart Rate Monitoring. A Practical Guide. Oxford University Press, Oxford (1993)Google Scholar
  8. 8.
    Bretscher, J., Saling, E.: pH values in the human fetus during labor. Am. J. Obstet. Gynecol. 97, 906–911 (1967)CrossRefGoogle Scholar
  9. 9.
    Tuffnell, D., Haw, W.L., Wilkinson, K.: How long does a fetal scalp blood sample take? BJOG 113, 332–334 (2006)CrossRefGoogle Scholar
  10. 10.
    Goldaber, K.G., Gilstrap, L.C., Leveno, K.J., Dags, J.S., McIntire, D.D.: Pathologic fetal acidemia. Obstet. Gynecol. 78, 1103–1107 (1991)Google Scholar
  11. 11.
    James, L.S., Weisbrot, I.M., Prince, C.E., Holaday, D.A., Apgar, V.: The acid-base status of human infants in relation to birth asphyxia and the onset of respiration. J. Paediatr. 52(4), 379–394 (1958)CrossRefGoogle Scholar
  12. 12.
    Westgren, M., Kuger, K., Ek, S., Grunevald, C., Kublickas, M., Naka, K., et al.: Lactate compared with pH analysis at fetal scalp blood sampling: a prospective randomised study. Br. J. Obstet. Gynaecol. 105, 29–33 (1998)CrossRefGoogle Scholar
  13. 13.
    Malin, G.L., Morris, R.K., Khan, K.S.: Strength of association between umbilical cord pH and perinatal and long-term outcomes: systematic review and meta-analysis. BMJ 340, c1471 (2010)CrossRefGoogle Scholar
  14. 14.
    ACOG Committee on Obstetric Practice: ACOG Committee Opinion No. 348, November 2006: Umbilical cord blood gas and acid-base analysis. Obstetrics and gynaecology, 108(5), p. 1319 (2006)Google Scholar
  15. 15.
  16. 16.
    Yeh, P., Emary, K., Impey, L.: The relationship between umbilical cord arterial pH and serious adverse neonatal outcome: analysis of 51 519 consecutive validated samples. BJOG: Int. J. Obstet. Gynaecol. 119(7), 824–831 (2012)CrossRefGoogle Scholar
  17. 17.
    Strachan, B.K., Sahota, D.S., Wijngaarden, W.J., James, D.K., Chang, A.M.: Computerised analysis of the fetal heart rate and relation to acidaemia at delivery. BJOG Int. J. Obstet. Gynaecol. 108(8), 848–852 (2001)CrossRefGoogle Scholar
  18. 18.
    Georgieva, A., Payne, S.J., Moulden, M., Redman, C.W.: Artificial neural networks applied to fetal monitoring in labour. Neural Comput. Appl. 22(1), 85–93 (2013)CrossRefGoogle Scholar
  19. 19.
    Jeżewski, M., Czabański, R., Wróbel, J., Horoba, K.: Analysis of extracted cardiotocographic signal features to improve automated prediction of fetal outcome. Biocybern. Biomed. Eng. 30(4), 29–47 (2010)Google Scholar
  20. 20.
    Keith, R.D., Beckley, S., Garibaldi, J.M., Westgate, J.A., Ifeachor, E.C., Greene, K.R.: A multicentre comparative study of 17 experts and an intelligent computer system for managing labour using the cardiotocogram. BJOG Int. J. Obstet. Gynaecol. 102(9), 688–700 (1995)CrossRefGoogle Scholar
  21. 21.
    Magenes, G., Signorini, M.G., Arduini, D.: Classification of cardiotocographic records by neural networks. In: Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks, IJCNN 2000, vol. 3, pp. 637–641. IEEE (2000)Google Scholar
  22. 22.
    Abdillah, A.A.: Suwarno: diagnosis of diabetes using support vector machines with radial basis function kernels. Int. J. Technol. 7(5), 849–858 (2016)CrossRefGoogle Scholar
  23. 23.
    Georgoulas, G., Stylios, C., Groumpos, P.: Feature extraction and classification of foetal heart rate using wavelet analysis and support vector machines. Int. J. Artif. Intell. Tools 15(03), 411–432 (2006)CrossRefGoogle Scholar
  24. 24.
    Georgoulas, G., Gavrilis, D., Tsoulos, I.G., Stylios, C., Bernardes, J., Groumpos, P.P.: Novel approach for fetal heart rate classification introducing grammatical evolution. Biomed. Signal Process. Control 2(2), 69–79 (2007)CrossRefGoogle Scholar
  25. 25.
    Warrick, P.A., Hamilton, E.F., Kearney, R.E., Precup, D.: Classification of normal and hypoxic fetuses using system identification from intrapartum cardiotocography. IEEE Trans. Biomed. Eng. 57, 771–779 (2010)CrossRefGoogle Scholar
  26. 26.
    Spilka, J., Chudáček, V., Koucký, M., Lhotská, L., Huptych, M., Janků, P., Georgoulas, G., Stylios, C.: Using nonlinear features for fetal heart rate classification. Biomed. Signal Process. Control 7(4), 350–357 (2012)CrossRefGoogle Scholar
  27. 27.
  28. 28.
    Friedman, J.H.: Greedy function approximation: a gradient boosting machine. Annals of statistics, pp. 1189–1232 (2001)Google Scholar
  29. 29.
    Hastie, T., Qian, J.: Glmnet Vignette (2014)Google Scholar
  30. 30.
    Altman, N.S.: An introduction to kernel and nearest-neighbour nonparametric regression. Am. Stat. 46(3), 175–185 (1992)Google Scholar
  31. 31.
    Liaw, A., Wiener, M.: Classification and regression by random Forest. R News 2(3), 18–22 (2002)Google Scholar
  32. 32.
    Breiman, L.: Random Forests. Statistics Department, University of California, Machine learning (2001)zbMATHGoogle Scholar
  33. 33.
    Bradley, A.P.: The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recogn. 30(7), 1145–1159 (1997)CrossRefGoogle Scholar
  34. 34.
    Witten, I.H., Frank, E.: Data Mining: Practical Machine Learning Tools and Techniques. Morgan Kaufmann, Burlington (2005)Google Scholar
  35. 35.
    Maaten, L.V.D., Hinton, G.: Visualizing data using t-SNE. J. Mach. Learn. Res. 9, 2579–2605 (2008)zbMATHGoogle Scholar
  36. 36.
    Noreen, E.W.: Computer intensive methods for hypothesis testing: An introduction (1989)Google Scholar
  37. 37.
    Sykes, G.S., Molloy, P.M., Johnson, P., Stirrat, G.M., Turnbull, A.C.: Fetal distress and the condition of newborn infants. Br. Med. J. (Clin. Res. Ed.) 287(6397), 943–945 (1983)CrossRefGoogle Scholar
  38. 38.
    Steer, P.J.: Fetal monitoring—Present and future. Eur. J. Obst. Gynecol. Reprod. Biol. 24(2), 112–117 (1987)CrossRefGoogle Scholar
  39. 39.
    Berg, P., Schmidt, S., Gesche, J., Saling, E.: Fetal distress and the condition of the newborn using cardiotocography and fetal blood analysis during labour. BJOG Int. J. Obst. Gynecol. 94(1), 72–75 (1987)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  • Rounaq Abbas
    • 1
  • Abir Jaafar Hussain
    • 1
    Email author
  • Dhiya Al-Jumeily
    • 1
  • Thar Baker
    • 1
  • Asad Khattak
    • 2
  1. 1.Department of Computer ScienceLiverpool John Moores UniversityLiverpoolUK
  2. 2.College of Technological InnovationZayed UniversityAbu DhabiUAE

Personalised recommendations