Author Login

Correspondence to Author: Juliana Medeiros Lima¹, Maria Victoria de Freitas², Marina Vellasco Oliveira Camelo de Castro³, Mariana Hamer Silva⁴, Catharine Harumi Konno⁵, Ana Lucia de Lima⁶, Vinícius Porto Alves⁷, Ana Karla de Sousa Batista⁸, Leocadia Felix de Araujo⁹, Thiago Augusto Rochetti Bezerra⁹. 

¹ Medical degree from Centro Universitário Lusíada. Medical Residency in Gynecology and Obstetrics at Hospital Municipal de Cubatão, São Paulo.
² Medical degree from Barão de Mauá University Center. Medical residency in Gynecology and Obstetrics at Guilherme Álvaro Hospital, Santos, São Paulo.
³ Medical degree from the University of Rio Verde, Aparecida de Goiânia Campus. Resident in Gynecology and Obstetrics at Guilherme Álvaro Hospital, Santos, São Paulo.
⁴ Medical degree from the Municipal University Center of Franca. Resident in Gynecology and Obstetrics at Guilherme Álvaro Hospital, Santos, São Paulo.
⁵ Medical degree from the Pontifical Catholic University of Paraná. Resident in Gynecology and Obstetrics at Guilherme Álvaro Hospital, Santos, São Paulo.
⁶ Medical degree from the Federal University of Roraima. Resident in Gynecology and Obstetrics at Jorge Rossmann Regional Hospital, Itanhaém, São Paulo.
⁷ Medical degree from Salvador University. Resident in Gynecology and Obstetrics at Jorge Rossmann Regional Hospital, Itanhaém, São Paulo.
⁸ Medical degree from the Federal University of Roraima. Resident in Gynecology and Obstetrics at Jorge Rossmann Regional Hospital, Itanhaém, São Paulo.
⁹ Medical degree from the Federal University of Fronteira Sul – UFFS. Passo Fundo, Rio Grande do Sul
¹º Medical student at the University of Ribeirão Preto, Guarujá Campus, São Paulo. Bachelor’s degree in Physical Education from the Federal University of São Carlos. Doctorate in Medical Sciences from the Ribeirão Preto School of Medicine, University of São Paulo, Ribeirão Preto, São Paulo.

DOI: 10.52338/jocogr.2025.5256

Abstract:

Artificial Intelligence (AI) has emerged as a promising tool for improving predictive capacity in obstetrics, enabling early detection of gestational complications with high potential for maternal and fetal morbidity and mortality. This study aimed to analyze, through a systematic review with meta-analysis, the role of AI in predicting obstetric complications, focusing on preeclampsia, preterm birth, postpartum hemorrhage, gestational diabetes mellitus, fetal distress, and neonatal mortality. The research was conducted according to the PRISMA protocol, covering the PubMed/ MEDLINE, Scopus, Web of Science, ScienceDirect, and SciELO databases, including studies published between 2019 and 2025. After screening and critical analysis, 47 studies were included in the qualitative synthesis and 21 in the meta-analysis. The results showed that deep learning models performed better, with a mean area under the curve (AUC) of 0.92 (95% CI: 0.88–0.95), followed by supervised machine learning algorithms (mean AUC of 0.86). The prediction of preeclampsia and gestational diabetes showed the best accuracy rates, while the outcomes of fetal distress and neonatal mortality exhibited greater heterogeneity (I² = 67%). Methodological analysis revealed that 68% of the studies had a low risk of bias, but there is still a lack of external validation and standardization of variables. It was concluded that AI showed high potential to revolutionize the screening and prediction of obstetric complications, optimizing prenatal care and clinical decision-making. However, the consolidation of these advances depends on the expansion of multicenter studies, the explainability of algorithms, and the ethical and safe integration of technology and medical practice. AI represents a milestone in predictive obstetric medicine, with a direct impact on reducing maternal and perinatal mortality.

Keywords: Artificial Intelligence; Obstetrics; Gestational complications; Machine learning; Risk prediction; Preeclampsia.

Citation:

Dr. Thiago Augusto Rochetti Bezerra, The Role Of Artificial Intelligence In Predicting Obstetric Complications. Journal of Clinical Obstetrics and Gynecology Research 2025.