An RDAU-NET model for lesion segmentation in breast ultrasound images

Abstract

Breast cancer is a common gynecological disease that poses a great threat to women health due to its high malignant rate. Breast cancer screening tests are used to find any warning signs or symptoms for early detection and currently, Ultrasound screening is the preferred method for breast cancer diagnosis. The localization and segmentation of the lesions in breast ultrasound (BUS) images are helpful for clinical diagnosis of the disease. In this paper, an RDAU-NET (Residual-Dilated-Attention-Gate-UNet) model is proposed and employed to segment the tumors in BUS images. The model is based on the conventional U-Net, but the plain neural units are replaced with residual units to enhance the edge information and overcome the network performance degradation problem associated with deep networks. To increase the receptive field and acquire more characteristic information, dilated convolutions were used to process the feature maps obtained from the encoder stages. The traditional cropping and copying between the encoder-decoder pipelines were replaced by the Attention Gate modules which enhanced the learning capabilities through suppression of background information. The model, when tested with BUS images with benign and malignant tumor presented excellent segmentation results as compared to other Deep Networks. A variety of quantitative indicators including Accuracy, Dice coefficient, AUC(Area-Under-Curve), Precision, Sensitivity, Specificity, Recall, F1score and M-IOU (Mean-Intersection-Over-Union) provided performances above 80%. The experimental results illustrate that the proposed RDAU-NET model can accurately segment breast lesions when compared to other deep learning models and thus has a good prospect for clinical diagnosis.

Lesion Segmentation in Breast Ultrasound Images using Optimized Marked Watershed Method

“…In terms of deep learning methods, many excellent image segmentation models have been gradually borrowed, improved and used. In the paper [22], several typical deep learning segmentation models are compared on dataset B. The results show that the RADU-NET model performs best.…”

“…An experienced radiologist sketched lesion boundary for each image as the ground-truth (GT). Dataset B is shared by the open source of the paper [22].…”

“…The supervised learning includes support vector machine (SVM), artificial neural network (ANN), Convolutional neural network (CNN), etc., which were applied to BUS images segmentation and made great progress ( [18], [19], [20], [21]. Zhuang proposed RDAU-NET model [22], which performs best on BUS images segmentation when compared to other modes. So far, deep learning models have been proven to be the best way to perform image segmentation.…”

Attention Gate ResU-Net for Automatic MRI Brain Tumor Segmentation

“…Besides, attention mechanism, which plays an important role in human perception, can effectively highlight the useful information while suppress the redundant one. Recently, attention mechanism has been receiving wide attention in a variety of computer tasks, such as natural language processing for machine translation [22], [23], natural image classification [24]- [26], salient object detection [27], [28], natural image segmentation [29]- [33] and video object segmentation [34] in computer vision fields, and medical image classification [35], [36] and medical image segmentation [37]- [41] in medical image analysis fields. There are many attempts that have embedded attention module into deep neural network architecture for improving the performance of image classification and image segmentation.…”

Stan: Small Tumor-Aware Network for Breast Ultrasound Image Segmentation

“…Zhuang el at. [14] proposed an RDAU-Net model, based on U-Net architecture, to perform the tumor segmentation task on BUS images, where dilated residual blocks and attention gates were used to replace the basic blocks and original skip connections in U-Net, respectively. Similarly, Hu et al [15] proposed a method that combined the dilated fully convolution network with a phase-based active contour model.…”