DADU: Dual Attention-based Deep Supervised UNet for Automated Semantic Segmentation of Cardiac Images

2025-04-18

AI Summary

This study addresses the challenges of low segmentation accuracy and ambiguous boundaries in automatic delineation of left/right ventricles and myocardial scar tissue from cardiac magnetic resonance (CMR) images. We propose an enhanced U-Net architecture integrating dual attention mechanisms and edge-guided deep supervision. Specifically, we design a novel channel-spatial collaborative attention module and incorporate Canny edge–driven skip connections to strengthen structural boundary awareness. Additionally, multi-level deep supervision loss (DS-Loss) is employed to mitigate gradient vanishing during training. Evaluated on public CMR datasets, our model achieves a Dice similarity coefficient of 98.0% and significantly reduces Hausdorff distance—particularly excelling in robust scar segmentation, a challenging small-target task. The proposed method consistently outperforms state-of-the-art approaches in both quantitative metrics and qualitative boundary fidelity.

Technology Category

Computer Vision: SegmentationMachine Learning: Calibration & Uncertainty QuantificationIntelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAGSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsGraph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

We propose an enhanced deep learning-based model for image segmentation of the left and right ventricles and myocardium scar tissue from cardiac magnetic resonance (CMR) images. The proposed technique integrates UNet, channel and spatial attention, edge-detection based skip-connection and deep supervised learning to improve the accuracy of the CMR image-segmentation. Images are processed using multiple channels to generate multiple feature-maps. We built a dual attention-based model to integrate channel and spatial attention. The use of extracted edges in skip connection improves the reconstructed images from feature-maps. The use of deep supervision reduces vanishing gradient problems inherent in classification based on deep neural networks. The algorithms for dual attention-based model, corresponding implementation and performance results are described. The performance results show that this approach has attained high accuracy: 98% Dice Similarity Score (DSC) and significantly lower Hausdorff Distance (HD). The performance results outperform other leading techniques both in DSC and HD.

Problem

• Automated segmentation of cardiac ventricles and scar tissue
• Improving accuracy in cardiac MRI image segmentation
• Addressing vanishing gradient in deep neural networks

Innovation

• Dual attention integrates channel and spatial features
• Edge-detection skip-connection enhances image reconstruction
• Deep supervision mitigates vanishing gradient issues