Neural network modeling in the prediction and prevention of perineal trauma in childbirth
AbstractThe aim of the study was to development a model for predicting perineal trauma in labor based on artificial intelligence and a comprehensive method of its prevention using non-drug methods.
Material and methods. The retrospective analysis of 994 delivery histories was performed to identify factors leading to birth traumatism. The data obtained became the basis for building the artificial neural network and developing the Childbirth prognostic computer program. At the second stage, a non-randomized open prospective comparative case-control study was conducted on data from 175 first-time pregnant women divided into two groups. The main group included 103 pregnant women who were prepared for childbirth according to an improved comprehensive program to prevent the perineal trauma, which included psychological training, training in perineal muscle tone control based on biofeedback using the Callibri Befit software package (Russia), and sessions of auriculomagnetopuncture. The reference group included 72 pregnant women who were prepared for childbirth according to the generally accepted program. Before the preventive program, all pregnant women were tested to determine the level of anxiety using the Spielberger-Hanin method. Statistical processing of the study results was carried out in Statistica 10 software package. Mean values of indicators and standard deviations were calculated. Student’s t-test (for independent and related data) was used to compare mean values. Clopper–Pearson confidence intervals were calculated for fractions. Nominal data were compared using Fisher’s exact test; relative risk score and its 95% confidence interval were calculated as a measure of the effect of factors. Differences were considered statistically significant at p<0.05.
Results. Increased perineal muscle tone was detected in women with high levels of reactive and personality anxiety, which was the basis for inclusion of this parameter in the neural network. The neural network report identified high risk of obstetric traumatism in 18 (17.5%) women of the main group and 9 (12.5%) of the reference group, medium risk in 21 (20.4%) and 26 (36.2%) women respectively. After the psychologist’s sessions in the main group, a statistically significant decrease in the level of anxiety (p<0.001) was noted in the patients. The number of deliveries complicated by soft tissue injuries of the birth canal and perineum in the main group was 8 (7.7%; 95% CI 0.029–0.126) of which 4 with high and 4 with average predicted risk, in the comparison group – 16 (22.2%; 95% CI 0.133–0.336): 9 and 7, respectively.
Conclusion. The Childbirth computer program developed on the basis of artificial intelligence allows the doctor to identify possible risks in advance, significantly improve the diagnostic system, conduct adequate preparation for pregnant women delivery and, thus, improve the quality of medical care. The data obtained demonstrate the high effectiveness of the developed psychophysical training program as a tool to reduce the incidence of birth traumatism.
Keywords:neural network modeling; birth trauma; psycho-emotional stress; biofeedback
Funding. The study had no sponsor support.
Conflict of interest. The authors declare no conflict of interest.
For citation: Zharkin N.A., Vassiliev P.M., Laipanova H.M., Miroshnikov A.E., Shatilova Iu.A., Kochetkov A.V. Neural network modeling in the prediction and prevention of perineal trauma in childbirth. Akusherstvo i ginekologiya: novosti, mneniya, obuchenie [Obstetrics and Gynecology: News, Opinions, Training]. 2024; 12. Supplement: 34–9. DOI: https://doi.org/10.33029/2303-9698-2024-12-suppl-34-39 (in Russian)
REFERENCES
1. Mamedova S.O., Kozhurova M.A., Kukharchik Yu.V., Raydyuk A.M. Perineal injuries in childbirth: risk factors. Smolenskiy meditsinskiy al’manakh [Smolensk Medical Almanac]. 2022; (1): 215–7. (in Russian)
2. Mikhel’son A.A., Mal’gina G.B., Luk’yanova K.D., Lazukina M.V., Lugovykh E.V., Varaksin A.N., et al. Early diagnosis and prevention of pelvic and urodynamic dysfunctions in women after delivery. Ginekologiya [Gynecology]. 2022; 24 (4): 295–301. (in Russian)
3. Sennikova Zh.V., Vorontsova N.A., Bakunovich M.P., Studenov G.V., Vekker I.R., Konstantinova O.D., et al. Perineal trauma in the obstetric hospital of the third level. Orenburgskiy meditsinskiy vestnik [Orenburg Medical Bulletin]. 2022; Х [4 (40)]: 28–32. (in Russian)
4. Durandin Yu.M., Marilova N.A., Toktar L.R. The state of connective tissue and its role in the pathogenesis of perineal trauma and genital prolapse. Vestnik Rossiyskogo universiteta druzhby narodov. Seriya: Meditsina [Bulletin of the Russian University of Peoples’ Friendship. Series: Medicine]. 2005; (4): 150–3. (in Russian)
5. Ishchenko A.I., Yashchuk A.G., Aleksandrov L.S., Chushkov Yu.V., Naftulovich R.A., Popova E.M., et al. Polyetiological pathogenesis of pelvic organ prolapse in women. Current understanding of the problem. Voprosy ginekologii, akusherstva i perinatologii [Problems of Gynecology, Obstetrics and Perinatology]. 2022; 21 (1): 76–84. DOI: https://doi.org/10.20953/1726-1678-2022-1-76-84 (in Russian)
6. Babich T.Yu., Plekho R.M., Yurchenko E.Yu., Suleymanova S.R., Grishko T.V. The use of fuctional biocontrol durihg the antenatal education of women. Mediko-farmatsevticheskiy zhurnal «Pul’s» [Medical and Pharmaceutical Journal «Pulse»]. 2020; 22 (3): 24–8. DOI: https://doi.org/10.26787/nydha-2686-6838-2020-22-3-24-28 (in Russian)
7. Prokhorova O.V. Preparation for the first childbirth: assessment of the impact on some characteristics of psychological status. Meditsinskaya nauka i obrazovaniye Urala [Medical Science and Education of the Urals]. 2020; 21 [3 (103)]: 17–20. DOI: https://doi.org/10.36361/1814-8999-2020-21-3-17-20 (in Russian)
8. Order of the Ministry of Health of Russia dated October 20, 2020 No. 1130n «On approval of the Procedure for the provision of medical care in the field of “obstetrics and gynecology|» (Registered with the Ministry of Justice of Russia on November 12, 2020 No. 60869 (in Russian)
9. Sugonyaev K.V. Is it possible to assess personality traits by means of Lüscher test? Psikhologiya. Psikhofiziologiya [Psychology. Psychophysiology]. 2019; 12 (2): 41–59. DOI: https://doi.org/10.14529/jpps190204 (in Russian)
10. Sergeev Yu.A., Sterleva E.A., Niazyan D.A. Application of neural networks in medicine. comparison of methods of neural network and group analysis of pathologies. StudNet. 2021; 4 (9). (in Russian)
11. Khasanov A.G., Shaybakov D.G., Zhernakov S.V., Men’shikov A.M., Badretdinova F.F., Sufiyarov I.F., et al. Neural networks in forecasting disease dynamics. Kreativnaya khirurgiya i onkologiya [Creative Surgery and Oncology]. 2020; 10 (3): 198–204. DOI: https://doi.org/10.24060/2076-3093-2020-10-3-198-204 (in Russian)