Computer aided decision support system to diagnose prenatal congenital heart defects from ultrasound images

Abstract

Congenital Heart Defect (CHD) is a birth defect formed during conception of the fetus. It is categorized into structural and functional heart defects. Approximately 33% to 50% of these defects are critical, requiring early intervention in the first year of neonatal life. Early diagnosis of prenatal CHD is a fundamental requirement in order to achieve a significant reduction in infant mortality rates. It is difficult for genuinely illustrating the hospital statistics for CHD in India, because countless of births happen at home in our nation with majority of it unsupervised by a qualified physician. Roughly 10% of present infant mortality in India is exclusively due to CHDs alone. In India Ultrasonography, Echocardiography and Color Doppler are being used as definitive tools for diagnosing prenatal CHD. The diverse merits of newlineUltrasound (US) modalities are the use of non-invasive imaging procedure with non-ionizing radiations, cost effectiveness, feature of providing pathologically significant diagnostic information and much more. The major drawback is the presence of inherent speckle noise, which makes the biological structures to appear with irregular boundaries. The major constraint in clinical radiology with manual investigation is the identification of diagnostically important sonographic markers. Diagnosing CHD from fetal heart 2 Dimensional (2D) US image requires unique set of progressive skills and exclusive exposure in anatomical structure identification of fetal heart. Owing to the nature of prenatal CHD diagnosis being a trivial and time consuming process, yet the diagnosis rate of prenatal CHD is low. newline newline