A review on identification of atrial septal defect using deep learning

Abstract

The third most prevalent kind of congenital cardiac disease is atrial septal defects (ASD). Even with extensive shunts, the majority of individuals remain asymptomatic throughout their infancy. Echocardiogram, Chest X-ray, Electrocardiogram (ECG), Cardiac catheterization, MRI, and CT scan may all be used to detect the abnormality. Deep learning can be employed for automated estimation of the defect from the test result. The goal of this review paper is first to provide an insight into ASD, the methods for diagnosis, the application of deep learning models for distinguishing the defect, defect detection accuracy and algorithm parameters.