Effective Attention-Guided Multi-Scale Medical Network for Skin Lesion Segmentation

TL;DR

This paper introduces a novel multi-scale attention-guided neural network architecture for skin lesion segmentation, combining multi-resolution fusion, dynamic attention, and an external attention bridge to improve accuracy and robustness.

Contribution

The paper presents a new encoder-decoder network with multi-scale residual structures, a Multi-Resolution Multi-Channel Fusion module, a Cross-Mix Attention Module, and an External Attention Bridge, advancing skin lesion segmentation.

Findings

Outperforms existing transformer and CNN-based models

Achieves higher segmentation accuracy on multiple datasets

Demonstrates robustness and effectiveness in challenging cases

Abstract

In the field of healthcare, precise skin lesion segmentation is crucial for the early detection and accurate diagnosis of skin diseases. Despite significant advances in deep learning for image processing, existing methods have yet to effectively address the challenges of irregular lesion shapes and low contrast. To address these issues, this paper proposes an innovative encoder-decoder network architecture based on multi-scale residual structures, capable of extracting rich feature information from different receptive fields to effectively identify lesion areas. By introducing a Multi-Resolution Multi-Channel Fusion (MRCF) module, our method captures cross-scale features, enhancing the clarity and accuracy of the extracted information. Furthermore, we propose a Cross-Mix Attention Module (CMAM), which redefines the attention scope and dynamically calculates weights across multiple contexts, thus improving the flexibility and depth of feature capture and enabling deeper exploration of subtle features. To overcome the information loss caused by skip connections in traditional U-Net, an External Attention Bridge (EAB) is introduced, facilitating the effective utilization of information in the decoder and compensating for the loss during upsampling. Extensive experimental evaluations on several skin lesion segmentation datasets demonstrate that the proposed model significantly outperforms existing transformer and convolutional neural network-based models, showcasing exceptional segmentation accuracy and robustness.