Can Algorithms Improve Birth? A Meta-Analysis of AI-Enhanced Digital Antenatal Care on Fetal Wellness

M. Ari Irawan; Fauzan Azmi Hasti Habibi Samosir; Muhammad Ikhwan; Fildza Sabhrina,D; Kaniya Atsilah Samosir

doi:10.32734/scripta.v8i1.25652

Authors

M. Ari Irawan RSAL Komang Makes Belawan, Medan, North Sumatera https://orcid.org/0009-0002-9324-1921
Fauzan Azmi Hasti Habibi Samosir Wates Husada Hospital, Gresik, East Java, Indonesia https://orcid.org/0009-0008-1639-6552
Muhammad Ikhwan Department of Obstetrics and Gynecology, Regina Marris Hospital, Medan, North Sumatera https://orcid.org/0009-0002-9324-1921
Fildza Sabhrina,D Wates Husada Hospital, Gresik, East Java, Indonesia https://orcid.org/0009-0007-9772-9613
Kaniya Atsilah Samosir RSAL Komang Makes Belawan, Medan, North Sumatera https://orcid.org/0009-0006-7212-7993

DOI:

https://doi.org/10.32734/scripta.v8i1.25652

Keywords:

Artificial intelligence, digital antenatal care, fetal growth restriction, fetal wellness, predictive analytics, analitik prediktif, kecerdasan buatan, kesejahteraan janin, pelayanan antenatal digital, pertumbuhan janin terhambat

Abstract

Background: Traditional digital antenatal care (ANC) fails to sufficiently lower global maternal and neonatal mortality. Artificial Intelligence (AI) shifts this paradigm by enabling proactive prevention through predictive analytics and real-time, personalized risk assessments.

Objectives: To systematically review and meta-analyze the impact of AI-enhanced digital monitoring and predictive alert systems on antenatal fetal wellness.

Methods: Following PRISMA guidelines, a systematic search of PubMed, Scopus, and Web of Science was conducted. Studies evaluating AI-enhanced digital monitoring or predictive algorithms in antenatal and intrapartum care were included to assess fetal growth restriction (FGR)-mortality, spontaneous delivery, and congenital defects.

Discussion: Ten studies were analyzed. AI-enhanced ANC demonstrated high diagnostic accuracy for FGR (AUC: 0.831–0.918) and outperformed standard cervical length assessment in predicting spontaneous preterm delivery (AUC 0.75 vs. 0.67; $p < 0.001$). AI-driven echocardiography boosted congenital heart defect detection sensitivity to 85% (vs. 34% clinically). Furthermore, graphical models revealed a 10-fold increased risk of perinatal morbidity when FGR combined with maternal diabetes (RR 9.8). AI also successfully identified delayed fetal brain maturation and predicted emergency Cesarean rates using heart rate variability. Overall, the models showed strong discriminative performance, yielding a pooled AUC of 0.858 (95% CI: 0.766–0.951).

Conclusion: AI-enhanced digital ANC significantly improves the early detection of critical fetal complications, substantially strengthening antenatal surveillance and intrauterine intervention management.

Keyword: Artificial intelligence, digital antenatal care, fetal growth restriction, fetal wellness, predictive analytics

Latar Belakang: Perawatan antenatal (Antenatal Care/ANC) digital tradisional belum cukup mampu menurunkan angka kematian maternal dan neonatal global. Kecerdasan Buatan (AI) mengubah paradigma ini dengan mengaktifkan pencegahan proaktif melalui analitik prediktif dan penilaian risiko yang dipersonalisasi secara real-time.

Tujuan: Meninjau secara sistematis dan melakukan meta-analisis terhadap dampak pemantauan digital berbasis AI dan sistem peringatan prediktif pada kesejahteraan janin selama masa antenatal.

Metode: Berdasarkan pedoman PRISMA, pencarian sistematis dilakukan pada database PubMed, Scopus, dan Web of Science. Studi yang dievaluasi mencakup pemantauan digital berbasis AI atau algoritma prediktif dalam perawatan antenatal dan intrapartum untuk menilai hambatan pertumbuhan janin (Fetal Growth Restriction/FGR)-mortalitas, persalinan spontan, dan cacat bawaan.

Pembahasan: Sepuluh studi dianalisis. ANC berbasis AI menunjukkan akurasi diagnostik yang tinggi untuk FGR (AUC: 0,831–0,918) dan mengungguli penilaian panjang serviks standar dalam memprediksi persalinan prematur spontan (AUC 0,75 vs. 0,67; $p < 0,001$). Ekokardiografi berbasis AI meningkatkan sensitivitas deteksi cacat jantung bawaan hingga 85% (dibandingkan standar klinis yang hanya 34%). Selain itu, model grafis menunjukkan peningkatan risiko morbiditas perinatal sebesar 10 kali lipat ketika FGR disertai dengan diabetes maternal (RR 9,8). AI juga berhasil mengidentifikasi keterlambatan pematangan otak janin dan memprediksi tingkat operasi sesar darurat menggunakan analisis variabilitas denyut jantung. Secara keseluruhan, model-model ini menunjukkan performa diskriminatif yang kuat, dengan nilai gabungan AUC sebesar 0,858 (95% CI: 0,766–0,951).

Kesimpulan: ANC digital berbasis AI secara signifikan meningkatkan deteksi dini komplikasi janin yang kritis, serta memperkuat pengawasan antenatal dan manajemen intervensi dalam rahim.

Kata kunci: Analitik prediktif, kecerdasan buatan, kesejahteraan janin, pelayanan antenatal digital, pertumbuhan janin terhambat