Abdominal Multi-organ Segmentation based on Multi-scale Convolutional Attention and Dynamic Feature Fusion

Abdominal multi-organ segmentation plays a critical role in medical image analysis, providing essential information and quantitative assessments for surgical navigation, organ transplantation, and radiation therapy planning. In recent years, deep learning-based automatic segmentation methods have emerged as a significant research focus, demonstrating remarkable improvements in segmentation speed, accuracy, and robustness while minimizing human errors. Among various deep learning architectures, U-Net has established itself as a predominant framework in medical image segmentation. In this study, we present an improved U-Net model that integrates a multi-scale convolutional attention mechanism and a dynamic feature fusion module, achieving efficient and accurate segmentation of abdominal multi-organ CT images. This advanced framework demonstrates significant clinical value by enhancing diagnostic accuracy and optimizing clinicians' workflow efficiency.