Paper Title
Efficient Epicardial Fat Segmentation in Cardiac CT Using Neural Networks

Abstract
In this paper, we present a epicardial fat segmentation methodology using the ConvNext model, a highly effici-ent neural network architecture with high performance in computer vision tasks. Ourmethodology uses 3D image segmentation to clearly outline epicardial fat tissue, an important component in the Cardiovascular system. By combining ConvNext's superior feature extraction capabilities with advanced segmentation techniques, we achieved significant improvements in both accuracy and computational efficiency over traditional methods. The proposed excellent performance on non-contrast CT datasets with an IOU of 0.9. Keywords – Neural Networks, Artificial Intelligence in Medical Imaging, Real Time Segmentation, Cardio Vascular diseases