In Psychiatry and clinical neurosciences
AIM : We applied natural language processing and machine learning to explore the disease-related language patterns which deserve objective measures for assessing language ability in Japanese patients with Alzheimer disease, while most previous studies have used large publicly available datasets in Euro-American languages.
METHODS : We obtained 276 speech-samples from 42 patients with Alzheimer disease and 52 healthy controls, aged 50 years or over. We used a natural language processing library for Python called spaCy with an add-on library called GiNZA, which is a Japanese parser based on Universal Dependencies designed to facilitate multilingual parser development. We used eXtreme Gradient Boosting for our classification algorithm. Each unit of part-of-speech and dependency was tagged and counted to create features such as tag-frequency and tag-to-tag transition-frequency. Each feature's importance was computed during the 100-fold repeated random sub-sampling validation and averaged.
RESULTS : The model resulted in an accuracy of 0.84 (SD = 0.06), and an Area Under the Curve of 0.90 (SD = 0.03). Among the features that were important for such predictions, 7 of the top 10 features were related to part-of-speech, while the remaining 3 were related to dependency. A box-plot analysis demonstrated that the appearance rates of content word-related features were lower among the patients, whereas those with stagnation-related features were higher.
CONCLUSION : This study demonstrated a promising level of accuracy for predicting Alzheimer disease and found the language patterns corresponding to the type of lexical-semantic decline known as "empty speech," which is regarded as a characteristic of Alzheimer disease. This article is protected by copyright. All rights reserved.
Momota Yuki, Liang Kuo-Ching, Horigome Toshiro, Kitazawa Momoko, Eguchi Yoko, Takamiya Akihiro, Goto Akiko, Mimura Masaru, Kishimoto Taishiro
2022-Dec-29
Alzheimer disease, Dementia, Machine Learning, Natural Language Processing, Speech-Language Pathology
