International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 11 | Nov 2025
p-ISSN: 2395-0072
www.irjet.net
FETAL BRAIN ABNORMALITIES USING YOLO FOR DETECTION AND SEGMENTATION Varsha J1, Vimala H S2 1Student , M.Tech IoT, Dept. of Computer Science, University of Visvesvaraya College of Engineering, Bengaluru,
India
2 Dr. Vimala H S, Professor, Dept. of Computer Science, University of Visvesvaraya College of Engineering,
Bengaluru, India -----------------------------------------------------------------------------***--------------------------------------------------------------------
Abstract - This study focuses on improving early
In this study, a new framework is presented that employs the You Only Look Once (YOLO) deep learning model to identify and categorize various fetal brain abnormalities. The approach is further strengthened by an integrated segmentation technique that accurately highlights and visualizes the specific brain regions affected, providing clearer insights for clinical evaluation.
identification and categorization of fetal brain abnormalities by applying deep learning methods, with an emphasis on the You Only Look Once (YOLO) architecture. The proposed system leverages YOLO’s fast and efficient object-detection capabilities to analyze fetal brain images acquired through modern imaging technologies. By training the model to automatically detect structural anomalies—such as ventriculomegaly, agenesis of the corpus callosum, and other developmental irregularities—the framework aims to enhance diagnostic precision and reduce the time required for assessment. The integration of YOLO enables real-time detection and classification, supporting quicker clinical decision-making and facilitating timely guidance for expecting families. Beyond detection and classification, the research incorporates a dedicated segmentation module designed to accurately outline the regions affected by the identified abnormalities. This segmentation component uses advanced image-processing techniques to generate detailed anatomical maps, improving the interpretability of the system’s predictions. Such visual outputs assist clinicians in understanding the extent and location of each abnormality, thereby supporting more targeted intervention strategies. By combining YOLO-based detection with high-resolution segmentation, the proposed approach offers a comprehensive framework for early and reliable evaluation of fetal brain anomalies, demonstrating the significant potential of deep learning in prenatal diagnosis and medical planning.
Fig-1: Use-Case Diagram for a Fetal Brain Abnormality Detection System
1.1 Motivation
Key Words: YOLO, fetal brain abnormalities, segmentation,
Early identification of fetal brain abnormalities is essential for ensuring timely medical care and reducing risks during pregnancy. Traditional diagnostic methods depend heavily on expert interpretation of ultrasound or MRI images, which can lead to variability and delayed detection. With the rapid growth of deep learning, there is a strong opportunity to enhance prenatal diagnosis through automated and reliable image analysis.
prenatal imaging, deep learning
1.INTRODUCTION Advances in prenatal healthcare have increasingly benefited from the use of artificial intelligence, particularly deep learning, to enhance the analysis of medical images. Detecting fetal brain abnormalities at an early stage is often difficult, making it essential to develop reliable tools that can improve diagnostic precision and simplify the classification of such conditions.
© 2025, IRJET
|
Impact Factor value: 8.315
YOLO’s real-time detection capabilities make it suitable for quickly identifying abnormalities with high accuracy. However, existing systems often lack precise localization and segmentation of affected regions.
|
ISO 9001:2008 Certified Journal
|
Page 808