RED-MAM: A residual encoder-decoder network based on multi-attention fusion for ultrasound image denoising.

• The proposed novel RED-MAM model for ultrasound image denoising, which can obtain richer features from images, and maintain the diversity of image feature information. • The proposed multi-attention fusion attention block is composed of multiple attention modules to focus on the importance of information on multiple feature domains by fusing different attention features. In the multi-attention fusion attention block, jump connections are used multiple times to avoid the loss of key feature information during reconstruction. • The denoising performance of the RED-MAM model is validated on the FH, HC18 and CAMUS ultrasound datasets. The result proves that the improved denoising model can effectively remove noise from ultrasound images. The proposed method has better performance on the fetal cardiac ultrasound dataset in noise suppression and structure preservation compared with other classic image denoising algorithms like BM3D, K-SVD, CNN10, etc. Compared with the state of art method RDN10, our method improved by 0.1 dB on the average peak signal-to-noise ratio (PSNR), improved by 0.5 dB compared with BM3D. The dependence of speckle noise in ultrasound images on image data makes the research of ultrasound image denoising tasks a great challenge. Several deep learning-based ultrasound image denoising methods have been proposed, but almost all of them suffer from the inability to focus on the importance of feature information on multiple domains at the same time. Therefore, this paper proposes a residual encoder-decoder based on multi-attention fusion attention module (RED-MAM) for ultrasound image denoising, which consists of five convolution layers, five deconvolution layers and two multi-attention fusion attention blocks. In order to make the denoising model better adapted to the properties of speckle noise, a multi-attention fusion attention block is constructed by several different attention modules. The multi-attention fusion attention block is introduced into the residual encoder-decoder denoising network to make the model focus on the ultrasound image texture structure information on multiple domains, thus enhancing the denoising effect of the model on ultrasound images. The performance of our proposed model is evaluated on three ultrasound image datasets. In terms of quantitative metrics, RED-MAM shows substantial improvement in Peak Signal-to-Noise Ratio (PSNR), Structural Similarity (SSIM) and Root Mean Square Error (RMSE). The qualitative results show that RED-MAM has significantly improved denoising performance and has good results in noise suppression as well as structure retention. Our method achieves state-of-the-art performance on all three ultrasound datasets. [ABSTRACT FROM AUTHOR]