Aisha Nurakhmet

"Machine Learning and Artificial Intelligence Approaches for Diagnosis of Cardiac diseases in Fetal and Pediatric Patients"

Started June 26, 2023

Abstract or project description

Abstract: Context: Diagnosing babies’ heart problems is tricky because their conditions can be complex and varied. Detecting these issues early and accurately is crucial for better treatment and outcomes. This research explores how machine learning and artificial intelligence can help doctors with diagnosing heart diseases in fetal and pediatric patients. Methods: In the course of the study, we review the literature on how AI/ML analyzes various types of data, including fetal echocardiograms (ultrasound of the heart in infants still in the womb), MRI of the heart in children and electronic medical records of children with confirmed heart problems. They use various ML/AI algorithms such as recurrent neural networks (RNNS), random forests and gradient boosting machines to extract important signs and classify patients into specific groups of heart diseases. The study evaluates the performance of the model using indicators such as sensitivity, specificity and AUC-ROC (a measure of model accuracy). Results: Not sure Conclusion: Using ML/AI to diagnose heart diseases in fetal and pediatric patients could be a game-changer. The models can detect intricate patterns in medical images and clinical information, leading to more precise and timely diagnoses. This integration of ML/AI in healthcare has the potential to transform how we care for young patients with heart issues, allowing for early treatments and personalized plans to improve their health and quality of life. However, before fully using these technologies, more research on larger datasets from multiple centers is needed. Medical experts, data scientists, and healthcare regulators must work together to ensure safe and responsible implementation in this special area of medicine.