AI in Healthcare

Qure.ai Is Building AI-Enabled Ultrasound Tools to Predict Preterm Birth in India — 12,000 Women, 1.8 Million Biospecimens, and the Country's Biggest Pregnancy Research Dataset

By Mustafa Shariff, Editor in Chief · June 03,2026 · 2 min read

Mustafa Shariff, Editor in Chief — Qure.ai Is Building AI-Enabled Ultrasound Tools to Predict Preterm Birth in India — 12,000 Women, 1.8 Million Biospecimens, and the Country's Biggest Pregnancy Research Dataset

Summary

  • Qure.ai + THSTI: Building India's Own AI Tool for Preterm Birth Risk Detection
  • Qure.ai has partnered with THSTI under the GARBH-INi initiative to develop AI ultrasound tools for preterm birth risk stratification. Here is the full story and why India-specific training data matters.

"For too long, our clinical tools relied on Western data that did not reflect Indian realities. GARBH-INi is helping bridge that gap with indigenous solutions that address the specific risk factors — malnutrition, low BMI, genetic markers — that drive preterm birth in India."

Digital health AI company Qure.ai has formalised a partnership with the Translational Health Science and Technology Institute under India's Department of Biotechnology to develop AI-enabled ultrasound tools for preterm birth risk detection and stratification — tools that will be trained on the largest pregnancy research dataset ever assembled in India, built specifically around Indian patient populations and Indian risk factors.

The partnership forms part of the GARBH-INi-AnandiMaa initiative — the commercialisation and clinical translation arm of the GARBH-INi research programme. GARBH-INi, short for Interdisciplinary Group for Advanced Research on Birth Outcomes, is India's largest pregnancy cohort study: 12,000 pregnant women enrolled across clinical sites, generating an accredited biorepository of more than 1.8 million biospecimens, over one million ultrasound images, and over 30,000 ultrasound videos. The dataset also incorporates microbiome analyses, genomic markers, maternal nutrition data, and longitudinal clinical follow-up across the full perinatal period.

Why India Needs Its Own Preterm Birth AI

India's preterm birth burden is enormous and specific. India accounts for an estimated 14% of global preterm births — approximately 3.5 million preterm deliveries annually — making it the country with the highest absolute number of preterm babies in the world. Preterm birth is the leading cause of neonatal mortality globally, contributing to 35% of all newborn deaths. In India, the consequences extend beyond neonatal survival: preterm birth is associated with long-term developmental disabilities, impaired cognitive development, and increased adult-onset disease risk that compounds India's non-communicable disease burden decades later.

The specific risk factors driving preterm birth in India differ materially from those in the Western populations on which most clinical prediction tools have been developed and validated. In India, low maternal body mass index and malnutrition — identified in the GARBH-INi cohort as primary modifiable risk factors — play a disproportionately large role. Microbiome dysbiosis patterns observed in Indian pregnant women differ from those in high-income country populations. Genetic markers relevant to preterm birth risk have different allele frequencies across India's diverse ethnic groups. And the clinical infrastructure for managing identified preterm birth risk — in terms of corticosteroid availability, neonatal care capability, and referral systems — varies dramatically across India's healthcare geography.

An AI tool trained on Western data and applied to Indian patients will inevitably misclassify risk in ways that are systematic rather than random: it will miss the specific Indian risk patterns it has not been trained on, and flag risk factors that are less prevalent in Indian populations. The GARBH-INi dataset's Indian origin — 12,000 Indian women, 1.8 million Indian biospecimens — makes it the correct foundation for a clinically valid Indian preterm birth prediction tool.

What Qure.ai Brings

Qure.ai — India's most globally recognised medical imaging AI company, with FDA-cleared products, CE marks, and deployment across hospitals in more than 50 countries — brings its deep learning ultrasound interpretation capabilities to the GARBH-INi dataset. The company's prior work on chest X-ray AI, head CT interpretation, and point-of-care ultrasound reporting demonstrates the technical capability to build clinically valid image-based AI from complex medical imaging datasets. The GARBH-INi dataset's over one million ultrasound images provides exactly the training volume that enables robust deep learning model development.

The specific AI outputs targeted include ultrasound-based cervical length measurement (the primary clinical predictor of preterm birth risk), uterine artery Doppler analysis, fetal biometry, and integrated risk stratification scores that combine imaging findings with clinical risk factors and biomarker data from the broader GARBH-INi dataset.

BHC Perspective

At Bengaluru Health Community, the Qure.ai-THSTI preterm birth AI partnership is one of the most clinically important stories in this edition. It demonstrates three things that matter: India is building AI healthcare tools on Indian data for Indian health problems; Qure.ai — a Bengaluru-headquartered company — is the technical partner of choice for the country's most important maternal health research programme; and the combination of world-class AI capability with world-class clinical research data creates solutions that are not just technically valid but clinically relevant for the patients they will serve.

Frequently Asked Questions

Q: What is the Qure.ai-THSTI preterm birth AI partnership?
A: Qure.ai has partnered with THSTI under the GARBH-INi-AnandiMaa initiative to develop AI-enabled ultrasound tools for preterm birth risk detection and stratification, trained on India's largest pregnancy research dataset — 12,000 women, 1.8 million biospecimens, 1+ million ultrasound images.

Q: Why does India need its own preterm birth AI?
A: India's preterm birth risk factors — low maternal BMI, malnutrition, specific microbiome patterns, Indian genetic markers — differ systematically from Western populations on which existing clinical tools are trained. AI trained on Indian data produces more clinically valid risk stratification for Indian patients.

Q: What is GARBH-INi?
A: India's largest pregnancy cohort study led by THSTI under the Department of Biotechnology — 12,000 enrolled pregnant women, an accredited biorepository of 1.8 million biospecimens and 1+ million ultrasound images, developing AI-based preterm birth prediction tools indigenous to India.