Artificial Intelligence in Obstetric Anomaly Scan: Heart and Brain.

Area of Science:

• Medical Imaging
• Artificial Intelligence in Medicine
• Prenatal Diagnostics

Background:

• Ultrasound is a primary tool for fetal evaluation but faces limitations like fetal position and maternal factors.
• Artificial intelligence (AI) shows promise in automating measurements, recognizing standard planes, and aiding disease diagnosis in fetal ultrasounds.
• The rise of electronic medical records and diagnostic imaging fuels AI's success in image recognition tasks.

Purpose of the Study:

• To review studies utilizing deep learning for ultrasound anomaly scan evaluation.
• Focus on complex fetal systems: heart and brain.
• Identify studies addressing the most frequent fetal anomalies.

Main Methods:

• Literature review of studies employing deep learning in fetal ultrasound.
• Analysis of AI applications in evaluating fetal heart and brain anomalies.
• Synthesis of findings on AI's role in anomaly detection.

Main Results:

• Deep learning algorithms demonstrate potential in enhancing the accuracy of fetal anomaly detection.
• AI assists in overcoming limitations of conventional ultrasound evaluations.
• Studies show AI's capability to reduce examination workload and duration.

Conclusions:

• Deep learning holds significant potential for improving prenatal diagnosis of fetal anomalies.
• AI-powered ultrasound anomaly scans can enhance diagnostic capabilities for fetal heart and brain abnormalities.
• Further research is warranted to integrate AI into routine obstetric practice for complex fetal evaluations.