Abstract
Purpose
Predicting individual risks for adverse outcomes in preterm infants is necessary for perinatal management and antenatal counseling for their parents. To evaluate whether a machine learning approach can improve the prediction of severe infant outcomes beyond the performance of conventional logistic models, and to identify maternal and fetal factors that largely contribute to these outcomes.
Methods
A population-based retrospective study was performed using clinical data of 31,157 infants born at < 32 weeks of gestation and weighing ≤ 1500 g, registered in the Neonatal Research Network of Japan between 2006 and 2015. We developed a conventional logistic model and 6 types of machine learning models based on 12 maternal and fetal factors. Discriminative ability was evaluated using the area under the receiver operating characteristic curves (AUROCs), and the importance of each factor in terms of its contribution to outcomes was evaluated using the SHAP (SHapley Additive exPlanations) value.
Results
The AUROCs of the most discriminative machine learning models were better than those of the conventional models for all outcomes. The AUROCs for in-hospital death and short-term adverse outcomes in the gradient boosting decision tree were significantly higher than those in the conventional model (p = 0.015 and p = 0.002, respectively). The SHAP value analyses showed that gestational age, birth weight, and antenatal corticosteroid treatment were the three most important factors associated with severe infant outcomes.
Conclusion
Machine learning models improve the prediction of severe infant outcomes. Moreover, the machine learning approach provides insight into the potential risk factors for severe infant outcomes.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Availability of data and material
Data that support the findings of this study are available from the corresponding author (TU) upon reasonable request and with permission from the Neonatal Research Network of Japan.
References
Chawanpaiboon S et al (2019) Global, regional, and national estimates of levels of preterm birth in 2014: a systematic review and modelling analysis. Lancet Glob Health 7(1):e37–e46
Moster D, Lie RT, Markestad T (2008) Long-term medical and social consequences of preterm birth. N Engl J Med 359(3):262–273
Barfield WD (2018) Public health implications of very preterm birth. Clin Perinatol 45(3):565–577
MacDorman MF, Gregory EC (2015) Fetal and perinatal mortality: United States, 2013. Natl Vital Stat Rep 64(8):1–24
Norman M et al (2019) Association between year of birth and 1-year survival among extremely preterm infants in Sweden during 2004–2007 and 2014–2016. JAMA 321(12):1188–1199
Inoue H et al (2017) Early mortality and morbidity in infants with birth weight of 500 grams or less in Japan. J Pediatr 190:112.e3-117.e3
Mactier H et al (2020) Perinatal management of extreme preterm birth before 27 weeks of gestation: a framework for practice. Arch Dis Child Fetal Neonatal Ed 105(3):232–239
David AL, Soe A (2018) Extreme prematurity and perinatal management. Obstet Gynaecol 20(2):109–117
Ushida T et al (2021) Antenatal prediction models for short- and medium-term outcomes in preterm infants. Acta Obstet Gynecol Scand 100(6):1089–1096
McLeod JS et al (2020) Comparing mortality risk models in VLBW and preterm infants: systematic review and meta-analysis. J Perinatol 40(5):695–703
Krittanawong C et al (2020) Machine learning prediction in cardiovascular diseases: a meta-analysis. Sci Rep 10(1):16057
Kourou K et al (2015) Machine learning applications in cancer prognosis and prediction. Comput Struct Biotechnol J 13:8–17
Wada K (2017) Our challenges for the intact survival of newborn babies. Perinatology 28(1):1–3
Kohavi R (1995) A study of cross-validation and bootstrap for accuracy estimation and model selection. Proc. International Joint Conference on AI, 1995, pp 1137–1145
Bergstra J et al (2015) Hyperopt: a Python library for model selection and hyperparameter optimization. Comput Sci Discov 8(1):014008
Hoerl AE, Kennard RW (1970) Ridge regression: biased estimation for nonorthogonal problems. Technometrics 12(1):55–67
Albawi S, Mohammed TA, Al-Zawi S (2017) Understanding of a convolutional neural network. In 2017 International Conference on Engineering and Technology (ICET)
Schmidhuber J (2015) Deep learning in neural networks: an overview. Neural Netw 61:85–117
Chawla NV (2002) SMOTE: synthetic minority over-sampling technique. J Artif Intell Res 16:321–357
Gholami R, Fakhari N (2017) Support vector machine: principles, parameters, and applications. In: Samui P, Sekhar S, Balas VE (eds) Handbook of neural computation. Academic, Netherlands, pp 515–535
Boulesteix A-L et al (2012) Overview of random forest methodology and practical guidance with emphasis on computational biology and bioinformatics. WIREs Data Min Knowl Discov 2(6):493–507
Schapire RE (1999) A brief introduction to boosting, in Proceedings of the 16th international joint conference on Artificial intelligence - Volume 2. Morgan Kaufmann Publishers Inc.: Stockholm, Sweden. pp 1401–1406
Friedman JH (2001) Greedy function approximation: a gradient boosting machine. Ann Statist 29(5):1189–1232
Murdoch WJ et al (2019) Definitions, methods, and applications in interpretable machine learning. Proc Natl Acad Sci 116(44):22071–22080
Lundberg S, Lee S-I (2017) A unified approach to interpreting model predictions in NIPS. 2017
Winter E (2002) The shapley value. Handbook of game theory with economic applications. Elsevier, Netherlands, pp 2025–2054
Podda M et al (2018) A machine learning approach to estimating preterm infants survival: development of the Preterm Infants Survival Assessment (PISA) predictor. Sci Rep 8(1):13743–13743
Ushida T et al (2020) Antenatal corticosteroids and outcomes in preterm twins. Obstet Gynecol 135(6):1387–1397
Carlo WA et al (2011) Association of antenatal corticosteroids with mortality and neurodevelopmental outcomes among infants born at 22 to 25 weeks’ gestation. JAMA 306(21):2348–2358
Melamed N et al (2016) The role of antenatal corticosteroids in twin pregnancies complicated by preterm birth. Am J Obstet Gynecol 215(4):482.e1–9
Norman M et al (2017) Association of short antenatal corticosteroid administration-to-birth intervals with survival and morbidity among very preterm infants: results from the EPICE cohort. JAMA Pediatr 171(7):678–686
Sun BZ et al (2020) Association of preeclampsia in term births with neurodevelopmental disorders in offspring. JAMA Psychiat 77(8):823–829
Nakamura N et al (2021) Mortality and neurological outcomes in extremely and very preterm infants born to mothers with hypertensive disorders of pregnancy. Sci Rep 11(1):1729
Xiao D et al (2018) Maternal chorioamnionitis and neurodevelopmental outcomes in preterm and very preterm neonates: a meta-analysis. PLoS ONE 13(12):e0208302
Boghossian NS et al (2018) Sex differences in mortality and morbidity of infants born at less than 30 weeks’ gestation. Pediatrics 142(6):e20182352
Acknowledgements
The authors thank Mr. Sota Suzuki for statistical analysis and Editage (www.editage.jp) for English-language editing.
Funding
This study was supported by a research grant from the Imai Seiichi Foundation and Kanzawa Medical Research Foundation awarded to TU.
Author information
Authors and Affiliations
Consortia
Contributions
TU, TK, and JB contributed to the concept and design of the study. TU and JB performed the statistical analyses. TU drafted the first version of the manuscript. TU, TK, YM, TN-K, YI, NN, MH, and HK were involved in analyzing and interpreting the data. TK and HK critically reviewed the manuscript, and all authors approved the final version of the manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors have no potential conflicts of interest to disclose.
Ethical approval
This study was approved by the Institutional Ethic Board of Nagoya University (approval number 2018-0026 on 9 May, 2018), and the use of this database was permitted by the Japan Neonatal Network Executive Committee.
Consent to participate
Informed consent was obtained from all parents at each facility.
Consent to publication
Not applicable.
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.
404_2022_6865_MOESM1_ESM.tif
Supplementary Figure 1. Histogram of the distribution of predicted probability and observed probability for each outcome. The distribution of the predicted probability and observed probability for each outcome are shown. IVH, intraventricular hemorrhage; PVL, periventricular leukomalacia. Supplementary file1 (TIF 35894 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.
About this article
Cite this article
Ushida, T., Kotani, T., Baba, J. et al. Antenatal prediction models for outcomes of extremely and very preterm infants based on machine learning. Arch Gynecol Obstet 308, 1755–1763 (2023). https://doi.org/10.1007/s00404-022-06865-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00404-022-06865-x