An intelligent clinical decision support system for predicting the outcome of an assisted reproductive technology protocol at various stages of its implementation

The article describes the logic of an intelligent clinical decision support system (CDSS), which is based on a set of machine learning models that allow predicting the outcome of an assisted reproductive technologies (ART) protocol at various stages of its implementation. To create all the prognostic models, data from the register of ART protocols, which enables tracing the influence of the woman's history and the course of the protocol on the health of the child from birth to three years of age, were used. The outcome of the ART protocol is expressed in the likelihood of pregnancy, the most common complications of its course, such as isthmic-cervical insufficiency, arterial hypertension, placenta previa, gestational diabetes mellitus, disturbances in the amount of amniotic fluid and premature rupture of the membranes, in a term and method of delivery, as well as in the state of health of the born child for three years. The impact of predicted pregnancy complications on the outcome of childbirth as well as the impact of predicted pregnancy complications, the date and method of delivery on the health of the born child, described in the health group and the predicted group of ICD-10 diagnoses, are taken into consideration. The CDSS is provided for in vitro fertilization protocols, including those using intracytoplasmic spermatozoa injection into the oocyte (IVF/ISKI) and cryotransfer. The CDSS contains 77 predictive models, of which 72 models are binary classifiers, 5 are regression models. Random Forest Algorithm was employed to create all machine learning models. The ROC-AUC value of the binary classifiers of the system is 0.936 95% CI [0.914; 0.958], the accuracy of binary classifiers is 0.897 95% CI [0.880; 0.915], F-test for regression models does not refute the model adequacy hypothesis. The application of such a system will make it possible to obtain an objective assessment drawing on a large amount of data, which is of particular interest for specialists in the field of ART, and to visually demonstrate to the clients of ART centers the main stages of the upcoming process.

1. Register of ART of the Russian Association of Human Reproduction. Access mode: http://rahr.ru/registr_otchet.php (accessed on 14.03.2022). (In Russ.)

2. Pessione F., De Mouzon J., Deveaux A., Epelboin S., Gervoise-Boyer M.-J., Jimenez C., Levy R., Valentin M., Viot G., Bergère M., Merlet F., Jonveaux P. Risques de morbidité maternelle et périnatale en fécodation in vitro: une étude nationale de cohorte française Gynécologie, obstétrique, fertilité & sénologie. 2020;48(4):351–358. Available at: https://www.sciencedirect.com/science/article/pii/S2468718920300519?via%3Dihub. (accessed on 15.03.2022). DOI: 10.1016/j.gofs.2020.02.002. (In French).

3. Sunderam S., Kissin D.M., Crawford S.B., Folger S.G., Jamieson D.J., Warner L., Barfield W.D. Assisted Reproductive Technology Surveillance – United States, 2014. MMWR Surveill Summ. 2017;66(6):1–24. Available at: https://www.cdc.gov/mmwr/volumes/66/ss/ss6606a1.htm (accessed on 15.03.2022). DOI: 10.15585/mmwr.ss6606a1.

4. Keshishian E.S., Tsaregorodtsev A.D., Ziborova M.I. The health status of children born after in vitro fertilization. Rossiyskiy Vestnik Perinatologii i Pediatrii = Russian Bulletin of Perinatology and Pediatrics. 2014;59(5):15–25. Available at: https://www.ped-perinatology.ru/jour/article/view/230?locale=ru_RU (accessed on 15.03.2022). (In Russ.)

5. Von Wolff M., Haaf T. In vitro fertilization technology and child health risks, mechanisms and possible consequences. Deutsches Ärzteblatt international. 2020;117(3):23–30. Available at: https://www.aerzteblatt.de/int/archive/article/211864 (accessed on 15.03.2022). DOI: 10.3238/arztebl.2020.0023.

6. Wennerholm U.B., Bergh C. Perinatal outcome in children born after assisted reproductive technologies. Upsala journal of medical sciences. 2020;125(2):158–166. Available at: https://ujms.net/index.php/ujms/article/view/5645 (accessed on 15.03.2022). DOI: 10.1080/03009734.2020.1726534.

7. McDonald S., Murphy K., Beyene J., Óhlsson A. Perinatal Outcomes of Singleton Pregnancies Achieved by In Vitro Fertilization: A Systematic Review and Meta-Analysis. Journal of obstetrics and gynaecology Canada. 2005;25(5):449–459. Available at: https://www.sciencedirect.com/science/article/abs/pii/S1701216316305278?via%3Dihub (accessed on 15.03.2022). DOI: 10.1016/S1701-2163(16)30527-8.

8. McDonald S., Murphy K., Óhlsson A. Perinatal outcomes of in vitro fertilization twins: a systematic review and meta-analyses. American Journal of Obstetrics & Gynecology. 2005;193(1):141–52. Available at: https://www.ajog.org/article/S0002-9378(04)02077-0/fulltext (accessed on 15.03.2022). DOI: 10.1016/j.ajog.2004.11.064.

9. Vaegter K.K., Lakic T.G., Olovsson M., Berglund L., Brodin T., Holte J. Which factors are most predictive for live birth after in vitro fertilization and intracytoplasmic sperm injection (IVF/ICSI) treatments? Analysis of 100 prospectively recorded variables in 8,400 IVF/ICSI single-embryo transfers. Fertility and Sterility. 2017;107(3):641–648.e2. Available at: https://www.fertstert.org/article/S0015-0282(16)63073-X/fulltext (accessed on 15.03.2022). DOI: 10.1016/j.fertnstert.2016.12.005.

10. Esteves S.C., Carvalho J.F., Bento F.C., Santos J. A Novel Predictive Model to Estimate the Number of Mature Oocytes Required for Obtaining at Least One Euploid Blastocyst for Transfer in Couples Undergoing in vitro Fertilization/Intracytoplasmic Sperm Injection: The ART Calculator. Frontiers in Endocrinology. 2019;10(99). Available at: https://www.frontiersin.org/articles/10.3389/fendo.2019.00099/full (accessed on 15.03.2022). DOI: 10.3389/fendo.2019.00099.

11. Ratna M.B, Bhattacharya S., Abdulrahim B., McLernon D.J. A systematic review of the quality of clinical prediction models in in vitro fertilization. Human Reproduction. 2020;35(1):100–116. Available at: https://academic.oup.com/humrep/article/35/1/100/5710852 (accessed on 15.03.2022). DOI: 10.1093/humrep/dez258.

12. Nelson S.M., Lawlor D.A. Predicting Live Birth, Preterm Delivery, and Low Birth Weight in Infants Born from In Vitro Fertilisation: A Prospective Study of 144,018 Treatment Cycles. PLOS Medicine. 2011;8(1): e1000386. Available at: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000386. (accessed on 15.03.2022). DOI: 10.1371/journal.pmed.1000386.

13. Pykhtina L.A., Filkina O.M., Gadzhimuradova N.D., Malyshkina A.I., Nazarov S.B. Risk factors and predicting health disorders in infants born from monocyesis after in vitro fertilization. Health Risk Analysis = Health Risk Analysis. 2017;1:56–65. Available at: https://journal.fcrisk.ru/eng/2017/1/7 (accessed on 15.03.2022). DOI: 10.21668/health.risk/2017.1.07.eng. (In Russ.)

14. Dukhovny D., Hwang S.S., Gopal D., Cabral H.J., Diop H., Stern J.E. Association of maternal fertility status and receipt of fertility treatment with healthcare utilization in infants up to age four. Journal of Perinatology. 2021;41(10):2408–2416. Available at: https://www.nature.com/articles/s41372-021-01003-y (accessed on 15.03.2022). DOI: 10.1038/s41372-021-01003-y.

15. Kovtun O.P., Plaxina A.N., Makutina V.A., Ankudinov N.O., Zilber N.A., Limanovskay O.V., Sinotova S.L. Information-analytical assessment systems for perinatal outcomes and children’s health status born by assisted reproductive technologies. Rossiyskiy Vestnik Perinatologii i Pediatrii = Russian Bulletin of Perinatology and Pediatrics. 2020;65(1):45–50. Available at: https://www.ped-perinatology.ru/jour/article/view/1056 (accessed on 15.03.2022). DOI: 10.21508/1027-4065-2020-65-1-45-50. (In Russ.)

16. Pedregosa et al. Scikit-learn: Machine Learning in Python. JMLR. 2011;12:2825–2830. Available at: https://www.jmlr.org/papers/volume12/pedregosa11a/pedregosa11a.pdf (accessed on 15.03.2022).

17. Breiman L. Random Forests. Machine Learning. 2001;45:5–32.

18. Korobov M. Morphological Analyzer and Generator for Russian and Ukrainian Languages. Analysis of Images, Social Networks and Texts. 2015:320–332. Available at: https://link.springer.com/chapter/10.1007/978-3-319-26123-2_31 (accessed on 15.03.2022). DOI: 10.1007/978-3-319-26123-2_31.

19. Kursa M.B., Rudnicki W.R. Feature Selection with the Boruta Package. Journal of Statistical Software. 2010;36(11):1–13. Available at: https://www.jstatsoft.org/article/view/v036i11 (accessed on 15.03.2022). DOI: 10.18637/jss.v036.i11.

20. He H., Bai Y., Garcia E.A., Li S. ADASYN: Adaptive synthetic sampling approach for imbalanced learning. Proceedings of the 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence), Hong Kong, China. 2008; 1322–1328. Available at: https://ieeexplore.ieee.org/document/4633969 (accessed on 15.03.2022). DOI: 10.1109/IJCNN.2008.4633969.

21. Lemaitre G., Nogueira F., Aridas C.K. Imbalanced-learn: Imbalanced-learn: A python toolbox to tackle the curse of imbalanced datasets in machine learning. JMLR. 2017;18(17):1–5. Available at: https://www.jmlr.org/papers/volume18/16-365/16-365.pdf (accessed on 15.03.2022).

22. Kohavi R. A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection. Proceedings of the Fourteenth International Joint Conference on Artificial Intelligence, San Mateo, CA. 1995;2(12):1137–1143.

23. Stone M. Cross-Validatory Choice and Assessment of Statistical Predictions. Journal of the Royal Statistical Society, Series B (Methodological). 1974;36(2):111–147. Available at: https://www.jstor.org/stable/2984809 (accessed on 15.03.2022).