ABSTRACT
Congenital heart defects (CHDs) is an interesting area of study of diagnosis the disease prenatally. There are 18 different CHDs, out of which few belong to the critical category. Our aim is to conduct a detailed review of the existing work in this area. Cardiac anomaly screening is the leading means of diagnosing congenital heart problems in utero. However, using ultrasound to diagnose these conditions is still difficult. The primary objective is to discuss various optimization methods that will aid in the diagnosis of prenatal CHD anomalies automatically without any human interference. The approach has been really transversal, with the use of image processing methods and artificial intelligence algorithms. Recent technological developments have made it possible for researchers to carry out computational intelligence techniques that would not have been conceivable using the traditional approaches. This chapter describes various preprocessing, segmentation, edge detection algorithms involved in this area of research. It also studies the different machine learning and deep algorithms actively involved in diagnosing prenatal CHD disease. This study is supported by our analysis of the benefits and drawbacks of current approaches as well as the potential future applications in this field.
