Khan Rashid, Zaman Asim, Chen Chao, Xiao Chuda, Zhong Wen, Liu Yang, Hassan Haseeb, Su Liyilei, Xie Weiguo, Kang Yan, Huang Bingding
College of Big Data and Internet, Shenzhen Technology University, Shenzhen, China.
College of Applied Sciences, Shenzhen University, Shenzhen, China.
Digit Health. 2024 Nov 18;10:20552076241291306. doi: 10.1177/20552076241291306. eCollection 2024 Jan-Dec.
The precise segmentation of kidneys from a 2D ultrasound (US) image is crucial for diagnosing and monitoring kidney diseases. However, achieving detailed segmentation is difficult due to US images' low signal-to-noise ratio and low-contrast object boundaries.
This paper presents an approach called deep supervised attention with multi-loss functions (MLAU-Net) for US segmentation. The MLAU-Net model combines the benefits of attention mechanisms and deep supervision to improve segmentation accuracy. The attention mechanism allows the model to selectively focus on relevant regions of the kidney and ignore irrelevant background information, while the deep supervision captures the high-dimensional structure of the kidney in US images.
We conducted experiments on two datasets to evaluate the MLAU-Net model's performance. The Wuerzburg Dynamic Kidney Ultrasound (WD-KUS) dataset with annotation contained kidney US images from 176 patients split into training and testing sets totaling 44,880. The Open Kidney Dataset's second dataset has over 500 B-mode abdominal US images. The proposed approach achieved the highest dice, accuracy, specificity, Hausdorff distance (HD95), recall, and Average Symmetric Surface Distance (ASSD) scores of 90.2%, 98.26%, 98.93%, 8.90 mm, 91.78%, and 2.87 mm, respectively, upon testing and comparison with state-of-the-art U-Net series segmentation frameworks, which demonstrates the potential clinical value of our work.
The proposed MLAU-Net model has the potential to be applied to other medical image segmentation tasks that face similar challenges of low signal-to-noise ratios and low-contrast object boundaries.
从二维超声（US）图像中精确分割肾脏对于肾脏疾病的诊断和监测至关重要。然而，由于超声图像的低信噪比和低对比度的目标边界，实现详细的分割很困难。
本文提出了一种用于超声分割的称为具有多损失函数的深度监督注意力（MLAU-Net）的方法。MLAU-Net模型结合了注意力机制和深度监督的优点来提高分割精度。注意力机制使模型能够有选择地关注肾脏的相关区域并忽略无关的背景信息，而深度监督则捕捉超声图像中肾脏的高维结构。
我们在两个数据集上进行了实验，以评估MLAU-Net模型的性能。带有注释的维尔茨堡动态肾脏超声（WD-KUS）数据集包含来自176名患者的肾脏超声图像，分为训练集和测试集，共计44,880张。开放肾脏数据集的第二个数据集有超过500张B模式腹部超声图像。在与最先进的U-Net系列分割框架进行测试和比较时，所提出的方法分别实现了最高的骰子系数、准确率、特异性、豪斯多夫距离（HD95）、召回率和平均对称表面距离（ASSD）分数，分别为90.2%、98.26%、98.93%、8.90毫米、91.78%和2.87毫米，这证明了我们工作的潜在临床价值。
所提出的MLAU-Net模型有可能应用于面临类似低信噪比和低对比度目标边界挑战的其他医学图像分割任务。
