A multicentric prospective cohort study to develop Artificial intelligence-assisted risk stratification tool for prediction of preterm birth.

Rationale/ gaps in existing knowledge - Preterm birth is a strong predictor of neonatal morbidity and mortality. It highlights the importance of identification of high-risk pregnancies and need for development of risk stratification tools to predict the occurrence of preterm births.

Novelty - Deep machine learning and artificial intelligence is the most recent technology and evolving nowadays. It is going to be a boom in health care system too. Development of artificial intelligence-assisted risk stratification tools to identify high-risk pregnancies and implementing these tools at community health centers will facilitate early diagnosis and timely referral, and reduce the magnitude of PTBs and other adverse obstetric outcomes too.

Objectives - The primary objective is to develop risk stratification tools for the prediction of preterm birth with the help of Artificial intelligence. And, secondary objectives are to estimate the prevalence of preterm birth in the study population and characteristics of pregnant women, risk factors, and causes of preterm births.

Methods - Approximately 5000 pregnant women will be recruited in the study to achieve a calculated sample size (preterm births) of 357 at all 4 study sites, and to be monitored till delivery. Once we finish the data collection, data will be analyzed with artificial intelligence and risk stratification tools for preterm birth prediction will be developed.

Expected outcome - The study is expected to identify major determinants of preterm birth and to develop risk stratification tools with the help of artificial intelligence. Implementation of these tools will lead to improved health facilities and better care for pregnant women and bring about major economical benefits. Later on, we will develop an App that would be easy to use even at a primary and community health care center. Early prediction of high-risk pregnancies and timely referral would reduce neonatal morbidity and mortality significantly.