ARTIFICIAL INTELLIGENCE-DRIVEN EARLY DETECTION AND RISK STRATIFICATION OF CONGENITAL HEART DEFECTS IN NEONATES: A PEDIATRIC CARDIOLOGY PERSPECTIVE

Abstract

Background: Congenital heart defects (CHDs) occur very commonly in newborns and, when untreated, can be very dangerous. Presently, echocardiography and pulse oximetry are the main ways doctors assess children in this field, but artificial intelligence (AI) is providing new ways to spot problems early and better classify risk in pediatric cardiology. There is not much information available about medical professionals’ awareness, acceptance and readiness for using AI.

Objective: The main purpose of this study is to learn what healthcare professionals think about using AI to identify and sort CHDs at birth. It looks at their knowledge of AI, benefits and concerns they might have and whether they are open to using such technologies in everyday neonatal care.

Methods: An organized online survey was set up and 120 people took part, consisting of pediatric cardiologists, neonatologists, general pediatricians and medical AI researchers. The questionnaire covered demographic information, CHD diagnosis procedures today, knowledge of AI, opinions on AI for help in diagnostics and roadblocks to using it. The quantitative data were provided with descriptive statistics, and the qualitative data were looked at using thematic analysis.

Results: Most of the participants agreed (over 60%) that AI can make both the accuracy and the speed of CHD diagnosis better. The majority of those surveyed outlined how AI can lead to more precise diagnoses, speed up the evaluation process and be a help where staff and resources are scarce. At the same time, issues regarding data security, clear ethics and untrained staff remained. Doctors were more likely to use AI if they had previous exposure, had organizational backing and access to results checking data.

Conclusion: The research points out that new AI techniques can make early detection and informed prediction of neonatal heart conditions more possible. The use of AI in healthcare may succeed, but the benefit depends on planned training, ethical rules and scientific proof. Such research should explore practical use, long-term benefits and work together with specialists from different fields to benefit pediatric cardiology with AI.