Abstract: The early detection of fetal brain abnormalities is critical for prompt intervention and better management of neonatal health. This project leverages the You Only Look Once (YOLO) algorithm, a state-of-the-art object detection technique, to achieve accurate and efficient detection of fetal brain anomalies from ultrasound images. Traditional methods of fetal brain analysis are often time-consuming and require specialized expertise, leading to potential delays in diagnosis. The YOLO algorithm, with its real-time processing capabilities, offers a promising solution by detecting multiple abnormalities in a single forward pass through the network.In this project, a comprehensive dataset of fetal ultrasound images is curated, encompassing a wide range of brain anomalies. The YOLO model is trained and fine-tuned to recognize specific patterns indicative of various conditions such as ventriculomegaly, holoprosencephaly, and others. The model's performance is evaluated based on metrics such as precision, recall, and mean Average Precision (mAP), ensuring robustness and reliability.

Index Terms: YOLO algorithm , Mean Average Precision (mAP) Object detection, Dataset, Training, Fine-tuning, Ultrasound images

| DOI: 10.17148/IARJSET.2024.11745