Researchers at Indian Institute of Technology Madras (IIT Madras) and Translational Health Science and Technology Institute (THSTI), Faridabad, have developed an India-specific AI model to determine the age of a foetus in a pregnant woman in the second and third trimesters precisely. This first-of a-kind research was carried out as a part of ‘Interdisciplinary Group for Advanced Research on Birth Outcomes – DBT India Initiative’ (GARBH-Ini) programme.

Accurate estimation of gestational age is fundamental for providing optimal care to pregnant women, monitoring foetal well-being, identifying and managing pregnancy complications and ensuring the best possible outcomes for both mother and baby.

In India, several challenges exist in accurately estimating the gestational age (GA) during pregnancy. This includes late initiation of antenatal care, typically as late as 14 weeks after gestation. Late initiation can make it challenging to accurately determine GA using traditional methods like the last menstrual period (LMP) due to uncertainties in LMP recall and irregular menstrual cycles.

The ethnic diversity is also an issue as India has a diverse population with variations in foetal growth patterns and biometric measurements compared to the populations for which standard GA estimation formulas were developed. Variability in foetal growth patterns influenced by factors such as maternal nutrition, health conditions and genetic factors can affect the accuracy of GA estimation. Therefore, population-specific models that account for these variations are essential for accurate GA assessment.

In many parts of the country, especially in rural and underserved areas, access to ultrasound facilities and trained sonographers are limited. This can hinder the use of ultrasound-based methods for accurate GA estimation during the second and third trimesters.

Also read
An aerial view of the incredible Grand Prismatic Spring in Wyoming, USA's Yellowstone National Park

Packed proteins for space

Every medical jotting makes a mighty ocean of clinical data that can yield life-saving insights

The healing power of data

AI that approximates the human brain 

Take the weight off: Making things lighter is the newest tool to fight climate change

The movement towards lightweight materials

The brain surviving long periods of oxygen deprivation is a big breakthrough, which opens possibilities of reviving dead people with all organs intact — such as those who drowned.

Bringing back the dead

India-specific Model

Currently, the age of a foetus is determined using a formula developed for the Western population. So, they are likely to be erroneous when applied to the Indian mothers due to variations in the growth of the foetus in the later part of pregnancy.

In light of this, a new model was developed using the data from the GARBH-Ini cohort study, which involved collecting detailed clinical data on pregnant women in India. Researchers from IIT-M and THSTI analysed this data to identify the key parameters and factors that influence GA estimation in the late trimesters. The resulting model — Garbhini-GA2 — takes into account the unique foetal biometry and growth patterns observed in the Indian population.

The Garbhini-GA2 model demonstrated higher accuracy in GA estimation compared to existing formulas like the Hadlock and InterGrowth-21st models.

Evaluation metrics such as root-mean-squared error, bias and pre-term birth rates were utilised to assess the performance of the model. It was found that the new model significantly reduced the median error in GA estimation by more than three times when compared to the Hadlock formula, indicating its better suitability for the Indian population.

GA estimation is vital for managing complications such as gestational diabetes, preeclampsia and other conditions that may require specific monitoring and treatment based on the stage of pregnancy. Hence, reliable GA data is essential for conducting research on pregnancy outcomes, including stillbirth, preterm birth and fetal growth restriction. Accurate GA estimation also ensures the correct classification of outcomes for epidemiological studies.

Welcoming this research, Dr Rajesh Gokhale, Secretary, Department of Biotechnology, Government of India, said, “The development of population-specific models for estimating gestational age is a commendable outcome of GARBH-Ini. These models are being validated across the country.”

This research was undertaken by Dr Himanshu Sinha, Associate Professor, Bhupat and Jyoti Mehta School of Biosciences, Department of Biotechnology, IIT Madras, Dr Shinjini Bhatnagar, the Principal Investigator of GARBH-Ini programme and a distinguished professor at THSTI.

Once validated, Garbhini-GA2 can be deployed in clinics across India, improving the care delivered by obstetricians and neonatologists, thereby reducing maternal and infant mortality rates in India.

comment COMMENT NOW