Abstract:

Aiming at the problems of computational redundancy and difficulty in segmenting fine structures of the original U-Net in medical image segmentation tasks， an improved U-Net algorithm based on attention mechanism and multi-scale fusion was proposed. Firstly， by integrating channel attention mechanism into the skip connections， the channels containing more important information were focused by the network， thereby reducing computational resource cost and improving computational efficiency. Secondly， the feature fusion strategy was added to increase the contextual information for the feature maps passed to the decoder， which realized the complementary and multiple utilization among the features. Finally， the joint optimization was performed by using Dice loss and binary cross entropy loss， so as to handle with the problem of dramatic oscillations of loss function that may occur in fine structure segmentation. Experimental validation results on Kvasir_seg and DRIVE datasets show that compared with the original U-Net algorithm， the proposed improved algorithm has the Dice coefficient increased by 1.82 and 0.82 percentage points， the SEnsitivity （SE） improved by 1.94 and 3.53 percentage points， and the Accuracy （Acc） increased by 1.62 and 0.04 percentage points， respectively. It can be seen that the proposed improved algorithm can enhance performance of the original U-Net for fine structure segmentation.

Key words: deep learning, medical image segmentation, U-Net, channel attention mechanism, multi-scale fusion

摘要：

针对原始U-Net在医学图像分割任务中计算冗余和难以划分细小结构等问题，提出一种基于注意力机制和多尺度融合的U-Net改进算法。首先，通过在跳跃路径上引入通道注意力机制，网络关注包含更重要信息的通道，从而减少计算资源开销，并提升计算效率；其次，增加特征融合策略为传递给解码器的特征图增加上下文信息，从而实现特征之间的互补和多重利用；最后，使用Dice损失和二元交叉熵损失进行联合优化，以应对细小结构分割时可能出现的损失函数剧烈振荡问题。在Kvasir_seg和DRIVE数据集上进行的实验验证的结果表明，与原始U-Net算法相比，所提改进算法的Dice系数分别提高了1.81和0.82个百分点，灵敏度（SE）分别提高了1.94和3.53个百分点，准确度（Acc）分别提高了1.62和0.04个百分点。可见，所提改进算法能够提升原始U-Net对于细小结构分割的性能。

关键词: 深度学习, 医学图像分割, U-Net, 通道注意力机制, 多尺度融合