CANES
PublicMedical image segmentation remains a crucial challenge in automated cardiac diagnosis. We present the CANES-Net architecture, a novel architecture combining U-Net with Transformer attention, Mamba selective state-space modeling with class-specific loss functions to improve multi-class cardiac structure segmentation on the ACDC dataset.
Creat：2025-06-11T08:26:39
Update：2025-06-11T08:38:45
1
Stars
0
Stars Increase