Abstract:Background Ultrasound imaging has been recognized as a powerful tool in clinical diagnosis. Nonetheless, the presence of speckle noise degrades the signal-to-noise of ultrasound images. Various denoising algorithms cannot fully reduce speckle noise and retain image features well for ultrasound imaging. The application of deep learning in ultrasound image denoising has attracted more and more attention in recent years. Methods In the article, we propose a generative adversarial network with residual dense connectivity and weighted joint loss (GAN-RW) to avoid the limitations of traditional image denoising algorithms and surpass the most advanced performance of ultrasound image denoising. The denoising network is based on U-Net architecture which includes four encoder and four decoder modules. Each of the encoder and decoder modules is replaced with residual dense connectivity and BN to remove speckle noise. The discriminator network applies a series of convolutional layers to identify differences between the translated images and the desired modality. In the training processes, we introduce a joint loss function consisting of a weighted sum of the L1 loss function, binary cross-entropy with a logit loss function and perceptual loss function. Results We split the experiments into two parts. First, experiments were performed on Berkeley segmentation (BSD68) datasets corrupted by a simulated speckle. Compared with the eight existing denoising algorithms, the GAN-RW achieved the most advanced despeckling performance in terms of the peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and subjective visual effect. When the noise level was 15, the average value of the GAN-RW increased by approximately 3.58% and 1.23% for PSNR and SSIM, respectively. When the noise level was 25, the average value of the GAN-RW increased by approximately 3.08% and 1.84% for PSNR and SSIM, respectively. When the noise level was 50, the average value of the GAN-RW increased by approximately 1.32% and 1.98% for PSNR and SSIM, respectively. Secondly, experiments were performed on the ultrasound images of lymph nodes, the foetal head, and the brachial plexus. The proposed method shows higher subjective visual effect when verifying on the ultrasound images. In the end, through statistical analysis, the GAN-RW achieved the highest mean rank in the Friedman test.

Deep Learning-Based Denoising of Acoustic Images Generated With Point Contact Method

Ultrasonic image denoising using machine learning in point contact excitation and detection method

Controllable Deep Learning Denoising Model for Ultrasound Images Using Synthetic Noisy Image.

A Universal Deep Learning Framework for Real-Time Denoising of Ultrasound Images

De-Noising of Photoacoustic Microscopy Images by Deep Learning

Deep Ultrasound Denoising Using Diffusion Probabilistic Models

Deep Ultrasound Denoising Without Clean Data

Denoising Plane Wave Ultrasound Images Using Diffusion Probabilistic Models

Deep sound-field denoiser: optically-measured sound-field denoising using deep neural network

Denoising Method for Photoacoustic Microscopy Using Deep Learning

Block-matching and 3D filtering-based denoising of acoustic images obtained through point contact excitation and detection method

Speckle Noise Reduction in Ultrasound Images using Denoising Auto-encoder with Skip Connection

Overview of Image Denoising Based on Deep Learning

Ultrasound image denoising using generative adversarial networks with residual dense connectivity and weighted joint loss

Addressing image and Poisson noise deconvolution problem using deep learning approaches

Unsupervised denoising of photoacoustic images based on the Noise2Noise network

Deep learning on image denoising: An overview

Deep Learning Denoising Applied to Regional Distance Seismic Data in Utah

De-Noising of Photoacoustic Microscopy Images by Attentive Generative Adversarial Network

Deep Learning-Based Denoising of Spatial Heterodyne Spectroscopy Interferograms Without Clear Images