Gurgaon: Doctors at six hospitals in India may soon start using an artificial intelligence tool, called Garbhini-GA2, to improve diagnosis of foetal abnormalities and get accurate gestational age for women who come for check-ups at later stages of their pregnancies.
Researchers from IIT Madras and Translational Health Science and Technology Institute (THSTI), who have developed the tool, say it will help reduce infant and maternal mortality rates, especially for women who often don’t find out about pregnancies till the second trimester. THSTI-Faridabad is an institute under the central govt’s department of biotechnology.
According to the National Family Health Survey, around 30-40 per cent of women who are pregnant come for their first scans in the second trimester, which means they don’t know their exact due date and doctors have to rely on dates of the last menstrual period to come up with an estimate.
Garbhini-GA2, which is installed into ultrasound machines, is meant to do away with this in accurately diagnosing gestational age even in scans carried out in the second trimester, its developers said.
“If you want to know the stage of pregnancy and due date, the most accurate estimate will come from the first trimester scan,” said Himanshu Sinha, associate professor at department of biotechnology in IIT-Madras.
“The tool’s primary function is to provide accurate pregnancy dating, which is crucial for managing pregnancy care. Accurate dating allows obstetricians to schedule necessary scans and tests at the right time, ensuring that potential complications are detected and managed effectively. It also helps in determining the optimal delivery window, thus reducing the risks associated with preterm and post-term births,” said Dr Shinjini Bhatnagar, principal investigator of the GARBH-Ini programme and professor at THSTI-Faridabad.
For now, researchers are in talks with six hospitals in NCR, Assam, Gujarat, Telangana and Puducherry. Researchers want to study the impact of the tool at six hospitals, and then roll it out across the country in collaboration with ultrasound manufacturers and hospitals.
“The team is currently focusing on smaller hospitals, aiming to minimise problems of adverse birth outcomes where gestational dating is a significant issue. The ultimate aim is to integrate it into ultrasound machines for effective clinical deployment”, said Dr Ramachandran Thiruvengadam, assistant professor at Pondicherry Institute of Medical Sciences, one of the collaborating institutions in the project.
The Garbh-Ini initiative, introduced by the central govt in May 2015 at the Civil Hospital in Gurgaon, initially involved a study of women who were followed from early pregnancy to delivery and post-partum. This cohort formed the foundation for Garbhini-GA2, which was developed by feeding troves of data to a machine learning tool that can read ultrasound scans.
