Comparative Study on Early detection of Congenital Heart Disease (CHD) using Fetal echocardiography

Abstract

This review explores the evolving world of detecting and handling heart conditions in babies
before they're born. It connects traditional ways of predicting a baby's outcome with the
exciting new tools of artificial intelligence (AI), along with new, simple tests that aren't
invasive.
Heart defects are a major health issue across the globe, accounting for nearly a third of all
serious birth defects. Despite this, routine screenings during pregnancy are still falling short,
often failing to catch these problems in non-specialist clinics. When a critical heart issue is
missed, it tragically leads to more newborns getting sick or dying.
For years, doctors have relied on a method called the "Cardiovascular Profile" (CVP) score to
check an unborn baby for heart failure and flag high-risk situations. A CVP score of 7 or lower
is a serious warning sign, strongly connected to a higher risk of death.
Now, AI offers a fresh approach to solve the problem of scans being dependent on a person’s
skill level. It can automatically analyze images and spot difficult patterns with a very high
accuracy rate (over 90%). Beyond that, pairing AI with a simple analysis of a mother's saliva
provides an accurate way to screen for specific, serious heart conditions without invasive
procedures.
However, before making this technology available in public, there are some big hurdles that
should be cleared. We need to test that all the data used is consistent, we have to get a clear
understanding of how the decisions are framed by AI (the "black box" problem), and we must
ensure it's rolled out in public in such a way that's fair and ethical for every patient.