WCSE 2022
ISBN: 978-981-18-3959-7 DOI: 10.18178/wcse.2022.06.041

Alzheimer's Disease Detection through Spontaneous Speech Using Attention Augmented Convolutional Neural Network

Jiyun Li, Ya Hai, Chen Qian

Abstract— Alzheimer's disease (AD) is a neurodegenerative disease which affects patients' thinking, mood, and memory. Once diagnosed, it cannot be cured or reversed. Mild cognitive impairment (MCI) is the early stage of Alzheimer's disease, and medication at this stage can be used to slow down or even stop its development. A large number of studies have shown that AD can cause language barriers. There are significant symptoms in language, which can be used for early detection of AD. In this paper, convolutional neural network (CNN) is applied to the early diagnosis of AD. In addition, we introduced Convolutional Block Attention Module (CBAM) and incorporate CBAM into the CNN architecture to enhance the performance of the model. Experimental results show that the proposed model in this paper achieves 84.87% and 83.00% classification accuracy in long speech tracks and short speech tracks of the Alzheimer's Disease Recognition Competition, improved 5.07% and 9.00% compared to the baseline system.

Index Terms—Alzheimer’s disease, cognitive decline detection, convolutional neural network, convolutional block attention module.

Jiyun Li, Ya Hai, Chen Qian
School of Computer Science and Technology Donghua University, Shanghai, CHINA

[Download]


Cite:Jiyun Li, Ya Hai, Chen Qian, "Alzheimer's Disease Detection through Spontaneous Speech Using Attention Augmented Convolutional Neural Network, " Proceedings of 2022 the 12th International Workshop on Computer Science and Engineering (WCSE 2022), pp. 285-290, June 24-27, 2022.