Research Progress and Clinical Implications of Generative Artificial Intelligence in Perinatal Health Care for Advanced Maternal Age Pregnant Women

Sep 24, 2025International journal of women's health

Generative Artificial Intelligence in Pregnancy Care for Older Mothers: Recent Advances and Clinical Uses

AI simplified

Circadian Biology on OpenScience ↗PubMed ↗DOI ↗OA ↗

Abstract

models achieve an AUC of 0.80-0.85 in predicting Group B Streptococcus infection.

Generative AI integrates multiple data sources for better healthcare efficiency.
Transformer models improve the accuracy of prenatal depression screening by 15-20% over traditional methods.
Challenges include data privacy risks, with 32% of maternal health institutions lacking encrypted data storage.
42% of clinicians report low trust in AI decision-making due to the 'black box' nature of models.
Only 18% of county-level hospitals utilize AI perinatal tools, highlighting urban-rural technological gaps.

AI simplified

OBJECTIVE: To analyze the current application status, technical characteristics, and challenges of Generative Artificial Intelligence () in perinatal health care for advanced maternal age pregnant women and explore targeted optimization strategies.

METHODS: A systematic literature review was conducted by searching PubMed, Web of Science, CNKI, and Wanfang Data from January 2020 to April 2025. Studies were included if they focused on Generative AI applications in perinatal care for women aged ≥35 years; 78 eligible studies (42 Chinese, 36 international) were finally included, covering technical applications, clinical validation, and ethical governance. We summarized the applications of Generative AI in risk prediction, personalized management, and remote monitoring, and analyzed issues related to data governance, technical limitations, resource allocation, and ethical supervision.

RESULTS: Generative AI improves healthcare efficiency by integrating multiple data sources for model construction, planning dynamic interventions, and facilitating remote monitoring. Specifically, GANs-based models achieve an AUC of 0.80-0.85 in predicting Group B Streptococcus infection, while Transformer models enhance the accuracy of prenatal depression screening by 15-20% compared to traditional methods. However, it faces challenges including data privacy risks (eg, 32% of maternal health institutions lack encrypted data storage), the "black box" nature of models (42% of clinicians report low trust in AI decision-making), urban-rural technological gaps (only 18% of county-level hospitals use AI perinatal tools), and ambiguous liability definitions.

CONCLUSION: Generative AI demonstrates significant application potential in perinatal care for advanced maternal age pregnant women. Promoting its implementation through technological innovation (eg, explainable AI), interpretability optimization, resource deployment (eg, lightweight mobile tools), and ethical supervision is crucial to improving maternal and infant health outcomes in China and globally.

Key numbers

0.80-0.85

AUC for Group B Streptococcus Prediction

AUC achieved by GANs-based models in clinical applications.

15-20%

Accuracy Improvement in Prenatal Depression Screening

Accuracy increase compared to traditional screening methods.

32%

Lack of Encrypted Data Storage

Percentage of institutions lacking necessary data security measures.

Full Text

We can’t show the full text here under this license. Use the link below to read it at the source.

View full text (PDF) ↗

Research Progress and Clinical Implications of Generative Artificial Intelligence in Perinatal Health Care for Advanced Maternal Age Pregnant Women

Abstract

Key numbers

Full Text

You found one interesting study. We’ll send the next 7.

what lands in your inbox each week:

Recent issues from the circadian biology brief