Artifact-Assisted multi-level and multi-scale feature fusion attention network for low-dose CT denoising.
Artifact-Assisted multi-level and multi-scale feature fusion attention network for low-dose CT denoising.
复制标题
DOI：
10.3233/xst-221149
复制
发表时间：
2022-06
影响因子：
3
通讯作者：
Xueying Cui;Yingting Guo;Xiong Zhong;Shangguan Hong;B. Liu;Anhong Wang
中科院分区：
文献类型：
--
作者：
Xueying Cui;Yingting Guo;Xiong Zhong;Shangguan Hong;B. Liu;Anhong Wang
背景与目的由于低剂量CT（LDCT）图像通常具有较高的噪声，这可能影响疾病诊断的准确性，本研究的目的是开发和评估一种新的伪影辅助特征融合注意力（AAFFA）网络，以提取和减少LDCT图像中的图像伪影和噪声。方法在AAFFA网络中构造特征融合注意块，进行局部多尺度伪影特征提取和由粗到细的渐进融合。提出了一种基于跳跃连接和注意力模块的多层次融合结构用于伪影特征提取。具体而言，长距离跳跃连接用于增强和融合具有不同深度水平的伪影特征。然后，融合的浅特征进入通道注意力以更好地提取伪影特征，并且融合的深特征被发送到像素注意力以聚焦伪影像素信息。最后，设计了伪影通道，提供丰富的伪影特征，指导噪声和伪影特征的提取。AAPM LDCT Challenge数据集用于训练和测试网络。使用视觉观察和定量指标，包括峰值信噪比（PSNR），结构相似性指数（SSIM）和视觉信息保真度（VIF）的性能进行评估。结果AAFFA网络使AAPM LDCT图像的平均PSNR/SSIM/VIF值分别从43.4961，0.9595，0.3926提高到48.2513，0.9859，0.4589。结论：提出的AAFFA网络能够有效地减少噪声和伪影，同时保持对象的边缘。视觉质量和量化指标的评估表明，与其他图像去噪方法相比，该方法有了进步。
BACKGROUND AND OBJECTIVE Since low-dose computed tomography (LDCT) images typically have higher noise that may affect accuracy of disease diagnosis, the objective of this study is to develop and evaluate a new artifact-assisted feature fusion attention (AAFFA) network to extract and reduce image artifact and noise in LDCT images. METHODS In AAFFA network, a feature fusion attention block is constructed for local multi-scale artifact feature extraction and progressive fusion from coarse to fine. A multi-level fusion architecture based on skip connection and attention modules is also introduced for artifact feature extraction. Specifically, long-range skip connections are used to enhance and fuse artifact features with different depth levels. Then, the fused shallower features enter channel attention for better extraction of artifact features, and the fused deeper features are sent into pixel attention for focusing on the artifact pixel information. Last, an artifact channel is designed to provide rich artifact features and guide the extraction of noise and artifact features. The AAPM LDCT Challenge dataset is used to train and test the network. The performance is evaluated using both visual observation and quantitative metrics including peak signal-noise-ratio (PSNR), structural similarity index (SSIM) and visual information fidelity (VIF). RESULTS Using AAFFA network improves the averaged PSNR/SSIM/VIF values of AAPM LDCT images from 43.4961, 0.9595, 0.3926 to 48.2513, 0.9859, 0.4589, respectively. CONCLUSIONS The proposed AAFFA network enables to effectively reduce noise and artifacts while preserving object edges. Assessment of visual quality and quantitative index demonstrates the progressive improvement compared with other image denoising methods.