Attention-based and micro designed EfficientNetB2 for diagnosis of Alzheimer's disease. (April 2023)

• Record Type: Journal Article
• Authors: Li, Hejie; Tan, Ying; Miao, Jiaqing; Liang, Ping; Gong, Jinnan; He, Hui; Jiao, Yuhong; Zhang, Fan; Xing, Yaolin; Wu, Donghan
• Is Part Of: Biomedical signal processing and control. Volume 82(2023)
• Publication Date: 2023-04
• Subjects: Alzheimer's disease -- Structural magnetic resonance imaging (sMRI) -- EfficientNetB2 -- Global attention mechanism -- Coordinate attention -- Micro design; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2023.104571
• Languages: English
• ISSNs: 1746-8094

Abstract

Abstract: Recently, many deep learning methods have been successfully used to diagnose Alzheimer's disease (AD) using brain imaging. However, structural magnetic resonance imaging (sMRI) of AD detects relatively small lesion areas, and it is difficult to distinguish the early lesions. These factors make it difficult to extract the key features in the subtle regions for discrimination in many studies. To address these issues, an attention-based and micro designed EfficientNetB2 (EB2) approach is proposed in this study to classify AD, mild cognitive impairment (MCI), and normal control (NC). First, the front portion of the EB2 uses a global attention mechanism (GAM) to improve the classification results by capturing significant features in three dimensions. Next, coordination attention (CA) was added to the EB2 model. The CA can automatically extract channel and location information from the sMRI two-dimensional slice data. The location information is critical for constructing spatial attention maps that help the model to identify the lesion areas accurately, thereby helping the model to extract features that are useful for classification. Finally, the model uses the micro design (MD) method of the ConvNeXt network, which investigates the effect of activation function and batch normalization layer on the model at the microscopic scale. MD can reduce the model complexity while improving classification ability considerably. The accuracy of the proposed method is 93.30%, 92.42% and 92.03% for AD/NC, AD/MCI and MCI/NC dichotomous data, respectively. Finally, the proposed method outperforms the existing convolutional neural networks, such as AlexNet, GoogleNet, MobileNetV2, MobileNetV3, and DenseNet. Highlights: Propose an attentional micro-design EB2 for Alzheimer's disease diagnosis. Utilize GAM retains more global lesion information in the front part of the model. CA helps the model in extracting the key feature information in the discriminant area. The MD mechanism can reduce the model complexity and improve its performance.