基于神经网络的多特征轻度认知功能障碍检测模型  

A neural network-based multi-feature detection model for mild cognitive impairment

作　　者：王欣 陈泽森 WANG Xin;CHEN Zesen(School of Foreign Languages,Sun Yat-sen University,Guangzhou 510275,China;School of Aeronautics and Astronautics,Sun Yat-sen University,Shenzhen 518107,China)

机构地区：[1]中山大学外国语学院,广东广州510275 [2]中山大学航空航天学院,广东深圳518107

出　　处：《中山大学学报（自然科学版）（中英文）》2023年第6期107-115,共9页Acta Scientiarum Naturalium Universitatis Sunyatseni

基　　金：教育部人文社会科学基金(22YJCZH179);中国科协科技智库青年人才计划(20220615ZZ07110400);中央高校基本科研业务费重点培育项目(23ptpy32)。

摘　　要：轻度认知功能障是介于正常衰老和老年痴呆之间的一种中间状态,是老年痴呆诊疗的关键阶段。因此,针对潜在MCI老年人群进行早期检测和干预,有望延缓语言认知障碍及老年痴呆的发生。本文利用患者在语言学表现变化明显的特点,提出了一种基于神经网络的多特征轻度认知障碍检测模型。在提取自然会话中的语言学特征的基础上,融合LDA模型的T-W矩阵与受试者资料等多特征信息,形成TextCNN网络的输入张量,构建基于语言学特征的神经网络检测模型。该模型在DementiaBank数据集上达到了0.93的准确率、1.00的灵敏度、0.8的特异度和0.9的精度,有效提高了利用自然会话对老年语言认知障碍检测的准确率。Mild cognitive impairment(MCI)is both an intermediate state between normal aging and Alzheimer's disease and the key stage in the diagnosis of Alzheimer's disease.Therefore,early detection and treatment for potential elderly can delay the occurrence of dementia.In this study,a neural network-based multi-feature detection model for mild cognitive impairment was proposed,which exploits the characteristics of patients with obvious changes in linguistic performance.The model is based on extracting the linguistic features in natural speech and integrating the T-W matrix of the LDA model with the subject data and other multi-feature information as the input tensor of the TextCNN network.It achieved an accuracy of 0.93,a sensitivity of 1.00,a specificity of 0.8,and a precision of 0.9 on the DementiaBank dataset,which effectively improved the accuracy of cognitive impairment detection in the elderly by using natural speech.

关 键 词：轻度认知功能障碍 自然会话 神经网络模型 多特征分析 会话分析 

分 类 号：H030[语言文字—语言学]