Multi-Organ Segmentation using Residual U-Net and tdistributed Stochastic Neighbor Embedding

Accurate organ segmentation in medical imaging plays a crucial role in diagnosis and treatment planning. This research presents a deep learning-based approach for multiorgan segmentation using a Residual U-Net architecture combined with t-distributed Stochastic Neighbor Embedding (t-SNE) for feature space visualization. The model is trained using annotated abdominal datasets to segment organs such as the large bowel, small bowel, and stomach. By integrating ResNet blocks into the U-Net encoder, the proposed model improves gradient flow, feature extraction, and segmentation accuracy. t-SNE visualization is utilized to interpret the clustering behavior of organ features in high-dimensional space. Experimental results indicate that the Residual U-Net outperforms the standard U-Net in terms of Dice Coefficient, and Hausdorff Distance. These findings highlight the potential of residual learning for precise and interpretable organ segmentation.