In obstetrics, artificial intelligence (AI) technology has been applied to ultrasound examination, fetal heart monitoring, MRI, and other areas to improve diagnostic efficiency and accuracy. Although there exists limitations and challenges associated with its application, AI demonstrates substantial potential in obstetrics. Looking forward, it will be essential for healthcare professionals to collaborate closely with AI technologies to improve the quality of obstetric care collectively. Therefor, this paper reviews the research progress of AI in obstetrics.

Objective:To investigate the impact of the number of cesarean deliveries on adverse maternal and neonatal outcomes.Methods:A retrospective analysis was conducted on 11 904 singleton pregnant women who underwent cesarean delivery at the Third Affiliated Hospital of Guangzhou Medical University from January 1st, 2019 to December 31st, 2023. The women were grouped according to the number of cesarean deliveries: those undergoing their first cesarean delivery (1CD group, 7 231 cases), those undergoing their second cesarean delivery (2CD group, 3 749 cases), those undergoing their third cesarean delivery (3CD group, 841 cases), and those undergoing their fourth or more cesarean deliveries (4CD group, 83 cases). Differences in clinical characteristics, related surgical procedures, and adverse maternal and neonatal outcomes among the groups were compared. Binary logistic regression analysis was used to assess the impact of the number of cesarean deliveries on related surgical procedures and adverse maternal and neonatal outcomes.Results:(1) During the 5-year period, the total number of women undergoing cesarean delivery in our hospital showed a slight downward trend, while the proportion of women undergoing three or more cesarean deliveries increased. (2) Compared with women undergoing their first cesarean delivery, women in each repeat cesarean delivery group were older, had higher proportions of advanced maternal age and pre-pregnancy body mass index, and had more pregnancies, deliveries, and induced abortions; the incidence of placenta previa, placental implantation, antepartum hemorrhage, gestational hyperglycemia, and failed trial of labor requiring conversion to surgery was higher, while the incidence of premature rupture of membranes was lower; the proportions of ureteral stent placement, adhesiolysis of the pelvic and abdominal cavities, uterine rupture, uterine reconstruction, uterine artery ligation, hysterectomy, postpartum hemorrhage, and postoperative intestinal obstruction were higher, and the amount of postpartum hemorrhage was greater; the gestational age at delivery of neonates was earlier, but the rates of preterm birth at 28-31 +6 and 32-33 +6 weeks of gestation were lower; the differences were statistically significant ( P<0.05) for all comparisons. (3) The number of cesarean deliveries was not an independent risk factor for the dose-dependent occurrence of placenta previa (a OR=0.99, 95% CI: 0.98-1.01; P=0.261). In women without placenta previa, the number of cesarean deliveries was not a risk factor for placental implantation (a OR=1.12, 95% CI: 0.90-1.39; P=0.320). However, in women with placenta previa, the number of cesarean deliveries was a risk factor for placental implantation (a OR=4.01, 95% CI: 3.08-5.22; P<0.001). In the overall population, the number of cesarean deliveries was a risk factor for ureteral stent placement, adhesiolysis of the pelvic and abdominal cavities, bladder rupture repair, uterine rupture, uterine reconstruction, uterine artery ligation, hysterectomy, postpartum hemorrhage, and preterm birth (all P<0.05). However, the number of cesarean deliveries was not a risk factor for postoperative intestinal obstruction, admission to the intensive care unit, neonatal asphyxia, admission to the neonatal intensive care unit, or neonatal death (all P<0.05). Conclusions:The number of cesarean deliveries could lead to adverse maternal and neonatal outcomes, but the relationship is not simply dose-dependent. It is speculated that the occurrence of severe adverse maternal and neonatal outcomes is more closely related to maternal complications and comorbidities, as well as whether multidisciplinary comprehensive management was received.

Artificial intelligence (AI), as a cutting-edge technology, is profoundly transforming obstetric practice. This review systematically summarizes how AI, through machine learning and deep learning algorithms, can efficiently process multi-source heterogeneous data, significantly improving the predictive accuracy of common complications such as preterm birth, gestational diabetes mellitus, and preeclampsia. It also facilitates the ultrasonographic recognition of fetal structural anomalies and automated biometric measurements, while providing more objective analysis of fetal heart rate monitoring. Concurrently, this article critically examines the challenges currently facing AI applications in obstetrics, including issues related to data quality and privacy protection, model interpretability, algorithmic fairness, accountability, and doctor-patient relationships, proposing corresponding strategies to address them. Future directions include multi-modal data fusion, the application of large language models, advancements in remote monitoring technologies, and interdisciplinary talent development. Despite numerous challenges, with ongoing technological evolution and improvements in ethical regulations, AI is expected to become deeply integrated into obstetric clinical workflows, offering sustained momentum for enhancing maternal and infant health outcomes.

Objective:To evaluate the clinical value of an obstetric intelligent assistant in predicting postpartum hemorrhage (PPH) after vaginal delivery.Methods:This retrospective cohort study included 4 832 women who delivered vaginally at ≥26 weeks of gestation at the Third Affiliated Hospital, Guangzhou Medical University between May 2023 and April 2025. Participants were categorized into PPH (382 cases, blood loss ≥500 ml within 24 h after delivery) and non-PPH groups (4 450 cases). Using traditional statistical methods combined with machine learning approaches, including support vector machines and extreme gradient boosting, supplemented with deep learning techniques, we developed a novel artificial neural network model—the obstetric intelligent assistant. This model provides a refined classification of PPH occurrence and estimated blood loss volume. The obstetric intelligent assistant integrates 70 antenatal and intrapartum risk factors through hospital information system interfacing to generate visualized risk probability outputs. Predictive performance was compared between the obstetric intelligent assistant and four conventional prediction tools (Chinese Labor Room Traffic Light System; Association of Women's Health, Obstetric and Neonatal Nurses; American College of Obstetrics and Gynecology Safe Motherhood Initiative; and California Maternal Quality Care Collaborative prediction tools) using receiver operating characteristic curve.Results:(1) For antenatal prediction, the obstetric intelligent assistant achieved an area under the curve of 0.826 (95% CI: 0.774-0.838), with sensitivity of 0.794 and specificity of 0.712, while the four conventional prediction tools showed area under the curve ranging from 0.569 to 0.586. (2) For intrapartum prediction, the obstetric intelligent assistant achieved an area under the curve of 0.786 (95% CI: 0.751-0.820), with sensitivity of 0.837 and specificity of 0.762, whereas the conventional tools showed area under the curve between 0.600 and 0.613. Conclusion:The obstetric intelligent assistant demonstrates superior performance in predicting PPH compared to conventional prediction tools.

Objective:To evaluate the effect of an obstetric artificial intelligence (AI) assistant combined with a family-centered health education model on maternal self-care ability, comfort status, and spousal caregiving ability.Methods:This prospective, single-center, parallel randomized controlled trial used 1∶1 randomization and was conducted as a superiority trial. Postpartum mothers and their spouses admitted to family-style single rooms at the Third Affiliated Hospital of Guangzhou Medical University between October 2024 and April 2025 were enrolled and randomly assigned to control or intervention groups using a random number table. The control group received conventional health education, while the intervention group received conventional health education plus the AI-assisted family-centered model. Interventions were administered at 2 hours, 6 hours, and 24 hours postpartum, and before discharge. Outcomes included maternal self-care ability, comfort status, and spousal caregiving ability, which were assessed at 2 hours postpartum and before discharge. Data were analyzed using independent and paired t-tests and Chi square tests. Results:Of the 88 mother-spouse dyads initially recruited, four were excluded due to mother-infant separation (e.g., neonatal jaundice), leaving 84 dyads (42 per group). After the intervention, the intervention group showed significantly higher maternal self-care ability scores [(192.81±13.80) vs. (181.00±21.41) scores, t=3.00], higher maternal comfort scores [(104.43±7.52) vs. (96.00±14.29) scores, t=3.38], and better spousal caregiving ability [(6.07±3.13) vs. (9.50±5.02) scores, t=－3.76] compared to the control group (all P<0.05). Conclusion:The obstetric AI assistant combined with a family-centered health education model significantly improved maternal self-care ability and comfort status, as well as spousal caregiving ability.

Objective:To explore the correlation between intrahepatic cholestasis of pregnancy(ICP)and ad-verse pregnancy outcomes.Methods:A total of 511 singleton pregnant women with ICP treated at The Third Affili-ated Hospital of Guangzhou Medical University from August 2017 to January 2024 were selected as the study sub-jects.Among them,patients were divided into the adverse pregnancy outcome group(n=49)and the control group without adverse pregnancy outcomes(n=462).The general and clinical data of the two groups were com-pared and analyzed.Results:①General situation:The number of pregnancies and deliveries,ICU transfer rate,total hospital stay,and total hospitalization costs were significantly higher in the adverse pregnancy outcome group compared to the control group(P<0.05).The number of prenatal check-ups,diagnostic gestational weeks,and gestational weeks at delivery were significantly lower compared to the control group(P<0.05).②Clinical symp-toms:The incidence of itching in the adverse pregnancy outcome group was lower compared to the control group(10.2%vs.26.6%,P<0.05),while other symptoms such as rash,fatigue,jaundice,and gastrointestinal symp-toms showed no significant difference between the two groups(P>0.05).③Laboratory examinations:Compared with the control group,patients in the adverse pregnancy outcome group had significantly the increased levels of alanine aminotransferase,aspartate aminotransferase,uric acid,urea nitrogen,and triglycerides,and significantly the decreased levels of alkaline phosphatase and fasting blood glucose,with statistical significance(P<0.05).Other biochemical indicators showed no significant difference between the two groups(P>0.05).④ICP grading and complications:The proportion of early-onset ICP,severe and very severe ICP in the adverse pregnancy out-come group was significantly higher compared to the control group(P<0.001);the proportion of adverse preg-nancy outcome group with pregnancy-induced hypertension was significantly higher compared to the control group;the incidence of preterm birth,fetal growth restriction,meconium-stained amniotic fluid,and fetal distress in the adverse pregnancy outcome group was significantly higher compared to the control group(P<0.001).⑤Neo-natal outcomes:The neonatal Apgar scores(1 min,5 min,10 min)and neonatal weight in the adverse pregnancy outcome group were lower compared to the control group(P<0.001),and the incidence of mild neonatal asphyx-ia was significantly higher,with a statistically significant difference(P<0.001).Conclusions:The severity of ICP is closely related to the occurrence of adverse pregnancy outcomes.Therefore,it is clinically necessary to pay at-tention to the grading of ICP,closely monitor the levels of total bile acids and liver enzymes,and try to avoid ad-verse pregnancy outcomes,especially intrauterine fetal death.

In obstetrics, artificial intelligence (AI) technology has been applied to ultrasound examination, fetal heart monitoring, MRI, and other areas to improve diagnostic efficiency and accuracy. Although there exists limitations and challenges associated with its application, AI demonstrates substantial potential in obstetrics. Looking forward, it will be essential for healthcare professionals to collaborate closely with AI technologies to improve the quality of obstetric care collectively. Therefor, this paper reviews the research progress of AI in obstetrics.

Objective:To investigate the impact of the number of cesarean deliveries on adverse maternal and neonatal outcomes.Methods:A retrospective analysis was conducted on 11 904 singleton pregnant women who underwent cesarean delivery at the Third Affiliated Hospital of Guangzhou Medical University from January 1st, 2019 to December 31st, 2023. The women were grouped according to the number of cesarean deliveries: those undergoing their first cesarean delivery (1CD group, 7 231 cases), those undergoing their second cesarean delivery (2CD group, 3 749 cases), those undergoing their third cesarean delivery (3CD group, 841 cases), and those undergoing their fourth or more cesarean deliveries (4CD group, 83 cases). Differences in clinical characteristics, related surgical procedures, and adverse maternal and neonatal outcomes among the groups were compared. Binary logistic regression analysis was used to assess the impact of the number of cesarean deliveries on related surgical procedures and adverse maternal and neonatal outcomes.Results:(1) During the 5-year period, the total number of women undergoing cesarean delivery in our hospital showed a slight downward trend, while the proportion of women undergoing three or more cesarean deliveries increased. (2) Compared with women undergoing their first cesarean delivery, women in each repeat cesarean delivery group were older, had higher proportions of advanced maternal age and pre-pregnancy body mass index, and had more pregnancies, deliveries, and induced abortions; the incidence of placenta previa, placental implantation, antepartum hemorrhage, gestational hyperglycemia, and failed trial of labor requiring conversion to surgery was higher, while the incidence of premature rupture of membranes was lower; the proportions of ureteral stent placement, adhesiolysis of the pelvic and abdominal cavities, uterine rupture, uterine reconstruction, uterine artery ligation, hysterectomy, postpartum hemorrhage, and postoperative intestinal obstruction were higher, and the amount of postpartum hemorrhage was greater; the gestational age at delivery of neonates was earlier, but the rates of preterm birth at 28-31 +6 and 32-33 +6 weeks of gestation were lower; the differences were statistically significant ( P<0.05) for all comparisons. (3) The number of cesarean deliveries was not an independent risk factor for the dose-dependent occurrence of placenta previa (a OR=0.99, 95% CI: 0.98-1.01; P=0.261). In women without placenta previa, the number of cesarean deliveries was not a risk factor for placental implantation (a OR=1.12, 95% CI: 0.90-1.39; P=0.320). However, in women with placenta previa, the number of cesarean deliveries was a risk factor for placental implantation (a OR=4.01, 95% CI: 3.08-5.22; P<0.001). In the overall population, the number of cesarean deliveries was a risk factor for ureteral stent placement, adhesiolysis of the pelvic and abdominal cavities, bladder rupture repair, uterine rupture, uterine reconstruction, uterine artery ligation, hysterectomy, postpartum hemorrhage, and preterm birth (all P<0.05). However, the number of cesarean deliveries was not a risk factor for postoperative intestinal obstruction, admission to the intensive care unit, neonatal asphyxia, admission to the neonatal intensive care unit, or neonatal death (all P<0.05). Conclusions:The number of cesarean deliveries could lead to adverse maternal and neonatal outcomes, but the relationship is not simply dose-dependent. It is speculated that the occurrence of severe adverse maternal and neonatal outcomes is more closely related to maternal complications and comorbidities, as well as whether multidisciplinary comprehensive management was received.

Objective:To explore the correlation between intrahepatic cholestasis of pregnancy(ICP)and ad-verse pregnancy outcomes.Methods:A total of 511 singleton pregnant women with ICP treated at The Third Affili-ated Hospital of Guangzhou Medical University from August 2017 to January 2024 were selected as the study sub-jects.Among them,patients were divided into the adverse pregnancy outcome group(n=49)and the control group without adverse pregnancy outcomes(n=462).The general and clinical data of the two groups were com-pared and analyzed.Results:①General situation:The number of pregnancies and deliveries,ICU transfer rate,total hospital stay,and total hospitalization costs were significantly higher in the adverse pregnancy outcome group compared to the control group(P<0.05).The number of prenatal check-ups,diagnostic gestational weeks,and gestational weeks at delivery were significantly lower compared to the control group(P<0.05).②Clinical symp-toms:The incidence of itching in the adverse pregnancy outcome group was lower compared to the control group(10.2%vs.26.6%,P<0.05),while other symptoms such as rash,fatigue,jaundice,and gastrointestinal symp-toms showed no significant difference between the two groups(P>0.05).③Laboratory examinations:Compared with the control group,patients in the adverse pregnancy outcome group had significantly the increased levels of alanine aminotransferase,aspartate aminotransferase,uric acid,urea nitrogen,and triglycerides,and significantly the decreased levels of alkaline phosphatase and fasting blood glucose,with statistical significance(P<0.05).Other biochemical indicators showed no significant difference between the two groups(P>0.05).④ICP grading and complications:The proportion of early-onset ICP,severe and very severe ICP in the adverse pregnancy out-come group was significantly higher compared to the control group(P<0.001);the proportion of adverse preg-nancy outcome group with pregnancy-induced hypertension was significantly higher compared to the control group;the incidence of preterm birth,fetal growth restriction,meconium-stained amniotic fluid,and fetal distress in the adverse pregnancy outcome group was significantly higher compared to the control group(P<0.001).⑤Neo-natal outcomes:The neonatal Apgar scores(1 min,5 min,10 min)and neonatal weight in the adverse pregnancy outcome group were lower compared to the control group(P<0.001),and the incidence of mild neonatal asphyx-ia was significantly higher,with a statistically significant difference(P<0.001).Conclusions:The severity of ICP is closely related to the occurrence of adverse pregnancy outcomes.Therefore,it is clinically necessary to pay at-tention to the grading of ICP,closely monitor the levels of total bile acids and liver enzymes,and try to avoid ad-verse pregnancy outcomes,especially intrauterine fetal death.

Artificial intelligence (AI), as a cutting-edge technology, is profoundly transforming obstetric practice. This review systematically summarizes how AI, through machine learning and deep learning algorithms, can efficiently process multi-source heterogeneous data, significantly improving the predictive accuracy of common complications such as preterm birth, gestational diabetes mellitus, and preeclampsia. It also facilitates the ultrasonographic recognition of fetal structural anomalies and automated biometric measurements, while providing more objective analysis of fetal heart rate monitoring. Concurrently, this article critically examines the challenges currently facing AI applications in obstetrics, including issues related to data quality and privacy protection, model interpretability, algorithmic fairness, accountability, and doctor-patient relationships, proposing corresponding strategies to address them. Future directions include multi-modal data fusion, the application of large language models, advancements in remote monitoring technologies, and interdisciplinary talent development. Despite numerous challenges, with ongoing technological evolution and improvements in ethical regulations, AI is expected to become deeply integrated into obstetric clinical workflows, offering sustained momentum for enhancing maternal and infant health outcomes.