RAR-U-Net: a Residual Encoder to Attention Decoder by Residual Connections Framework for Spine Segmentation under Noisy Labels

TL;DR

This paper introduces RAR-U-Net, a novel deep learning framework for spine segmentation that effectively handles noisy labels through residual connections, attention modules, and an adaptive denoising strategy, demonstrating competitive results.

Contribution

The paper presents a new residual encoder-decoder architecture with attention modules and an adaptive denoising strategy for improved medical image segmentation under noisy labels.

Findings

Achieved competitive performance on spine CT segmentation benchmark.

Effectively mitigated the impact of noisy labels during training.

Enhanced feature transfer and refinement through residual and attention mechanisms.

Abstract

Segmentation algorithms for medical images are widely studied for various clinical and research purposes. In this paper, we propose a new and efficient method for medical image segmentation under noisy labels. The method operates under a deep learning paradigm, incorporating four novel contributions. Firstly, a residual interconnection is explored in different scale encoders to transfer gradient information efficiently. Secondly, four copy-and-crop connections are replaced by residual-block-based concatenation to alleviate the disparity between encoders and decoders. Thirdly, convolutional attention modules for feature refinement are studied on all scale decoders. Finally, an adaptive denoising learning strategy (ADL) is introduced into the training process to avoid too much influence from the noisy labels. Experimental results are illustrated on a publicly available benchmark database of spine CTs. Our proposed method achieves competitive performance against other state-of-the-art methods over a variety of different evaluation measures.