Computer Engineering and Applications ›› 2024, Vol. 60 ›› Issue (5): 139-145.DOI: 10.3778/j.issn.1002-8331.2211-0150

• Pattern Recognition and Artificial Intelligence • Previous Articles     Next Articles

Personalized Dynamically Ensemble for Alzheimer’s Disease Auxiliary Diagnostics Model

LIANG Haolin, PAN Dan, ZENG An, YANG Baoyao, Xiaowei Song   

  1. 1.School of Computers, Guangdong University of Technology, Guangzhou 510006, China
    2.School of Electronics and Information, Guangdong Polytechnic Normal University, Guanzhou 510540, China
    3.Clinical Research Centre, Surrey Memorial Hospital, Surrey V3V 1Z2, Canada
  • Online:2024-03-01 Published:2024-03-01

个性化动态集成的阿尔茨海默症辅助诊断模型

梁浩霖,潘丹,曾安,杨宝瑶,Xiaowei Song   

  1. 1.广东工业大学 计算机学院,广州 510006
    2.广东技术师范大学 电子与信息学院,广州 510540
    3.素里纪念医院 临床研究中心,不列颠哥伦比亚省 V3V 1Z2

Abstract: Aiming at the problem that most of the Alzheimer’s disease (AD) classification models do not develop specific strategies for input samples, resulting in the easy neglect of personalized differential information between samples, a novel AD classification model, namely personalized dynamically ensemble convolution neural network (PDECNN), is proposed. Considering the difference in degeneration degree of brain regions between input samples, PDECNN involves an attention-net to evaluate the degeneration degree of each brain region specific to the input sample. Based on the estimated results of the attention-net, a dynamic ensemble strategy is newly designed to select and fuse brain region features for AD identification. In addition, by redesigning the loss function, the problem that the optimal gradient of unselected brain regions cannot be obtained is solved, thus improving the AD classification performance. The experimental results show that compared with AD classification models, the classification accuracy of PDECNN in the AD vs. HC (healthy cognition), MCIc (mild cognitive impairment who will convert to AD) vs. HC, and MCIc vs. MCInc (mild cognitive impairment who will not convert to AD) experiments can be increased by 4%, 11%, and 8%, respectively. The experimental results also find that the degenerate brain regions identified by the PDECNN correlate with AD’s clinical manifestations.

Key words: Alzheimer’s disease (AD), dynamic ensemble strategy, ensemble learning, convolutional neural network

摘要: 针对阿尔茨海默症(AD)分类模型大多没有针对输入样本制定特定的策略,导致容易忽略样本间的个性化差异信息的问题,提出个性化动态集成AD分类模型。该模型考虑到输入样本间脑区退化程度的差异性,利用注意力机制评估特定于输入样本的各脑区退化程度,并根据脑区退化程度对脑区特征进行挑选和融合;同时通过重新设计损失函数,解决未被选中脑区无法获得优化梯度的问题,从而提高AD分类性能。实验结果表明,该模型在AD vs.HC(正常组)、MCIc(会向AD转化的轻度认知障碍)vs.HC以及MCIc vs.MCInc(不会向AD转化的轻度认知障碍)中的分类准确率表现分别提升4%、11%以及8%。同时,模型定位到的退化脑区功能与AD临床表现具有高度一致性。

关键词: 阿尔茨海默症(AD), 动态集成策略, 集成学习, 卷积神经网络