U-Net-based Models for Skin Lesion Segmentation: More Attention and Augmentation

TL;DR

This study evaluates various U-Net-based models with attention mechanisms and data augmentation techniques for skin lesion segmentation, demonstrating improved accuracy and robustness in early melanoma detection.

Contribution

The paper introduces a novel combination of attention modules in U-Net architectures and compares multiple augmentation strategies, advancing skin lesion segmentation methods.

Findings

U-Net-Resnet50 and R2U-Net achieved top performance metrics.

Attention modules like CBAM and AG improve segmentation with minimal computational overhead.

Sequential use of pyramid, AG, and CBAM blocks outperforms individual applications.

Abstract

According to WHO[1], since the 1970s, diagnosis of melanoma skin cancer has been more frequent. However, if detected early, the 5-year survival rate for melanoma can increase to 99 percent. In this regard, skin lesion segmentation can be pivotal in monitoring and treatment planning. In this work, ten models and four augmentation configurations are trained on the ISIC 2016 dataset. The performance and overfitting are compared utilizing five metrics. Our results show that the U-Net-Resnet50 and the R2U-Net have the highest metrics value, along with two data augmentation scenarios. We also investigate CBAM and AG blocks in the U-Net architecture, which enhances segmentation performance at a meager computational cost. In addition, we propose using pyramid, AG, and CBAM blocks in a sequence, which significantly surpasses the results of using the two individually. Finally, our experiments show that models that have exploited attention modules successfully overcome common skin lesion segmentation problems. Lastly, in the spirit of reproducible research, we implement models and codes publicly available.