First Affiliated Hospital of Ningbo University, Ningbo, 315020, China; Faculty of Electrical Engineering and Computer Science, Ningbo University, Ningbo 315210, China.
Faculty of Electrical Engineering and Computer Science, Ningbo University, Ningbo 315210, China; Zhejiang Key Laboratory of Mobile Network Application Technology, Ningbo University, Ningbo 315210, China; Zhejiang Engineering Research Center of Advanced Mass Spectrometry and Clinical Application, Ningbo University, Ningbo 315210, China.
Comput Biol Med. 2023 Oct;165:107401. doi: 10.1016/j.compbiomed.2023.107401. Epub 2023 Aug 30.
Identifying progressive mild cognitive impairment (pMCI) and stable mild cognitive impairment (sMCI) plays a significant role in early Alzheimer's disease (AD) diagnosis, which can effectively boost the life quality of patients. Recently, convolutional neural network (CNN)- based methods using structural magnetic resonance imaging (sMRI) images have shown effective for AD identification. However, these CNN-based methods fail to effectively explore the feature extraction of disease-related multi-scale tissues, such as ventricles, hippocampi and cerebral cortex. To address this issue, we propose an end-to-end disease-related attentional UNet framework (DAUF) for identifying pMCI and sMCI, by embedding a devised dual disease-related attention module (DAM) and a novel tree-structured feature fusion classifier (TFFC). Specifically, DAM leverages the complementarity between feature maps and attention maps and the complementary features from the encoder and decoder, so as to highlight discriminative semantic and detailed features. Additionally, TFFC is a powerfully joint multi-scale feature fusion and classification head, by employing the homogeneity among multi-scale features, so that the discriminative features of the multi-scale tissues are adequately fused for enhancing classification performance. Finally, extensive experiments demonstrate the superior performance of DAUF, with the effectiveness of DAM and TFFC on identifying pMCI and sMCI subjects.
识别进展性轻度认知障碍（pMCI）和稳定轻度认知障碍（sMCI）在早期阿尔茨海默病（AD）诊断中起着重要作用，可以有效提高患者的生活质量。最近，基于卷积神经网络（CNN）的方法使用结构磁共振成像（sMRI）图像已被证明对 AD 识别有效。然而，这些基于 CNN 的方法未能有效探索与疾病相关的多尺度组织（如脑室、海马体和大脑皮层）的特征提取。为了解决这个问题，我们提出了一种用于识别 pMCI 和 sMCI 的端到端疾病相关注意力 U 型网络框架（DAUF），通过嵌入一种设计的双疾病相关注意力模块（DAM）和一种新颖的树状结构特征融合分类器（TFFC）。具体来说，DAM 利用特征图和注意力图之间的互补性以及来自编码器和解码器的互补特征，以突出有区别的语义和详细特征。此外，TFFC 是一种强大的联合多尺度特征融合和分类头，通过利用多尺度特征之间的同质性，使多尺度组织的判别特征充分融合，以提高分类性能。最后，广泛的实验证明了 DAUF 的优越性能，DAM 和 TFFC 对识别 pMCI 和 sMCI 受试者的有效性。
