Deep Learning Model to Predict In-hospital Mortality of Newborns during Congenital Heart Disease Surgery

Many parents are concerned about the cost of saving their child's life. The operation's cost depends on the pathology's nature and the chosen clinic's class. The human body functions as a single system where each organ performs its function. The heart is the main organ of the circulatory system and is responsible for filling all the blood vessels in the body. Surgery in 72% of diseases gives a chance for a complete recovery of the child. Its success depends on timing.

In this paper, an AI-induced deep learning model has been proposed to predict in-hospital mortality of newborns in congenital heart disease surgery. If the structure of the heart chambers or large vessels is different from normal, this indicates a defect. Heart disease is a disease caused by changes in the structure of valves, septa or blood vessels. These defects can lead to poor blood circulation in the body and depending on the affected area. Almost all heart defects are curable, often with surgery. Modern medicine has many successful cases of surgical treatment of heart defects in adults and children.

The proposed model reached 68.41% of training accuracy and 84.83% of testing accuracy, 83.44% training false discovery rate and 85.18% testing false discovery rate, 78.48% training false omission rate and 84.72% testing false omission rate, 70.26% training Positive likelihood ratio and 82.40% of testing positive likelihood ratio and 80.15% of training negative likelihood ratio and 82.97% of testing negative likelihood ratio.

With the development of modern surgery, early correction of CHD is possible even in low birth weight and premature babies. During surgery, the heart and lungs are cut off from the bloodstream, during which it is enriched with oxygen, which is distributed throughout the body. If the case is complicated, additional surgery may be required over a period of several months to 1 year from the previous surgery.