Network and Data Security Key Laboratory of Sichuan Province, University of Electronic Science and Technology of China, Chengdu, Sichuan, China; Sichuan Industrial Internet Intelligent Monitoring and Application Engineering Research Center, Chengdu University of Technology, Chengdu, China.
Network and Data Security Key Laboratory of Sichuan Province, University of Electronic Science and Technology of China, Chengdu, Sichuan, China.
Comput Biol Med. 2024 Sep;179:108847. doi: 10.1016/j.compbiomed.2024.108847. Epub 2024 Jul 15.
The UNet architecture, which is widely used for biomedical image segmentation, has limitations like blurred feature maps and over- or under-segmented regions. To overcome these limitations, we propose a novel network architecture called MACCoM (Multiple Attention and Convolutional Cross-Mixer) - an end-to-end depthwise encoder-decoder fully convolutional network designed for binary and multi-class biomedical image segmentation built upon deeperUNet. We proposed a multi-scope attention module (MSAM) that allows the model to attend to diverse scale features, preserving fine details and high-level semantic information thus useful at the encoder-decoder connection. As the depth increases, our proposed spatial multi-head attention (SMA) is added to facilitate inter-layer communication and information exchange, enabling the network to effectively capture long-range dependencies and global context. MACCoM is also equipped with a convolutional cross-mixer we proposed to enhance the feature extraction capability of the model. By incorporating these modules, we effectively combine semantically similar features and reduce artifacts during the early stages of training. Experimental results on 4 biomedical datasets crafted from 3 datasets of varying modalities consistently demonstrate that MACCoM outperforms or matches state-of-the-art baselines in the segmentation tasks. With Breast Ultrasound Image (BUSI), MACCoM recorded 99.06 % Jaccard, 77.58 % Dice, and 93.92 % Accuracy, while recording 99.50 %, 98.44 %, and 99.29 % respectively for Jaccard, Dice, and Accuracy on the Chest X-ray (CXR) images used. The Jaccard, Dice, and Accuracy for the High-Resolution Fundus (HRF) images are 95.77 %, 74.35 %, and 95.95 % respectively. The findings here highlight MACCoM's effectiveness in improving segmentation performance and its valuable potential in image analysis.
UNet 架构广泛应用于生物医学图像分割，但存在特征图模糊、过分割或欠分割区域等问题。为了克服这些局限性，我们提出了一种新的网络架构，称为 MACCoM（多注意力和卷积交叉混合器），这是一种端到端的深度编码器-解码器全卷积网络，专为二进制和多类生物医学图像分割而设计，构建在更深的 UNet 之上。我们提出了一种多范围注意力模块（MSAM），使模型能够关注不同尺度的特征，保留精细的细节和高级语义信息，因此在编码器-解码器连接中很有用。随着深度的增加，我们提出了空间多头注意力（SMA），以促进层间通信和信息交换，使网络能够有效地捕获长距离依赖关系和全局上下文。MACCoM 还配备了我们提出的卷积交叉混合器，以增强模型的特征提取能力。通过整合这些模块，我们在训练的早期阶段有效地结合了语义相似的特征，并减少了伪影。在从 3 个不同模态的数据集构建的 4 个生物医学数据集上进行的实验结果表明，MACCoM 在分割任务中的表现优于或与最先进的基线相匹配。在乳房超声图像（BUSI）上，MACCoM 记录了 99.06%的 Jaccard、77.58%的 Dice 和 93.92%的准确率，而在胸部 X 射线（CXR）图像上，记录了 99.50%、98.44%和 99.29%的 Jaccard、Dice 和准确率。在高分辨率眼底（HRF）图像上，Jaccard、Dice 和准确率分别为 95.77%、74.35%和 95.95%。这些发现强调了 MACCoM 在提高分割性能方面的有效性及其在图像分析中的宝贵潜力。
