基于改进EfficientNet的阿尔兹海默症图像分类  

Alzheimer’s Disease Image Classification Based on Improved EfficientNet

作　　者：朱剑波 葛明锋 董文飞[2] ZHU Jian-bo;GE Ming-feng;DONG Wen-fei(School of Intelligence and Information Engineering,Shandong University of Traditional Chinese Medicine,Jinan 250355,China;Suzhou Institute of Biomedical Engineering and Technology,Chinese Academy of Sciences,Suzhou 215163,China)

机构地区：[1]山东中医药大学智能与信息工程学院,山东济南250355 [2]中国科学院苏州生物医学工程技术研究所医学检验技术研究室,江苏苏州215163

出　　处：《计算机与现代化》2023年第6期56-61,共6页Computer and Modernization

基　　金：国家重点研发计划项目(2021YFB3600117);中科院仪器设备研制项目(YJKYYQ20200038);江苏省重点研发计划项目(BE2019683)。

摘　　要：为提升卷积神经网络用于阿尔兹海默症MRI图像分类的效果,提出一种融合自适应注意力机制和数据增强技术的卷积神经网络FAMENET。通过引入数据增强技术和Focal Loss损失函数缓解数据不平衡现象;重构优化主干网络Effi⁃cientNet,在保持精度的情况下减少模型参数量和网络的计算量;引入自适应注意力机制,解决输入图片进行特征提取下采样过程导致的信息丢失问题。在公开数据集进行大量对比实验,FAMENET的分类准确率达到79.95%,AUC值达到82.54%,设计的消融实验也充分证明了所提出的各个模块和网络的有效性。To improve the effectiveness of the convolutional neural network for Alzheimer’s disease MRI image classification,a convolutional neural network FAMENET is proposed,which integrates an adaptive attention mechanism and data enhancement technique to alleviate data imbalance by introducing a data augmentation technique and Focal Loss loss function.The network is reconfigured to reduce the number of model parameters and the computational effort of the network while maintaining accuracy.The adaptive attention mechanism is introduced to solve the information loss problem caused by the downsampling of input im⁃ages for feature extraction.In a large number of comparative experiments on public datasets,the classification accuracy of FAMENET reaches 79.95%and the AUC value reaches 82.54%.The designed ablation experiments also fully demonstrate the ef⁃fectiveness of the proposed modules and networks.

关 键 词：卷积神经网络 阿尔兹海默症 自适应注意力机制 数据增强 医学图像分类 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]