基於語料的失智患者辨識系統

隨著高齡化社會的來臨，失智患者越來越成為一個需要重視的課題。而失智患者的篩檢，現階段主要是依靠醫師問診、認知功能評斷，腦部影像以及抽血檢查來完成。在未來失智症患者可能增多的前提下，我們希望借助機器學習與深度學習的技術，達成認知功能評斷的任務輔助。本論文使用的數據，主要來自臺北市立聯合醫院仁愛院區仁鶴軒以及臺大醫院記憶門診，被診斷為失智患者以及非失智患者(即正常人)，對於「偷餅乾」圖片的獨白文本。我們預期：失智患者因為一些腦內功能的缺失，影響語言的表達能力、獨白內容的流暢度以及深度。本研究會從最基礎的機器學習分類，乃至深度學習模型分類進行分析。分析過程分為兩個階段，第一個部分是透過語言學的方式擷取特徵，希望能藉由詞性分布、以及與詞性相關的特徵進行分類；第二部份則透過詞嵌入的方式，加入上下文的考量，並導入BERT的相關技術，從中找出能有效分辨失智患者的演算法。

關鍵字

失智症；特徵工程；自然語言處理；機器學習；深度學習

並列摘要

With the advent of an aging society, dementia has become a more and more important issue. Nowadays, the detection of dementia mainly relies on physician consultation, the test of cognitive function, fMRI images, and blood sampling. Under the premise that we estimate that the number of patients with dementia may increase in the future, we hope to use machine learning and deep learning techniques to achieve task assistance for the test of cognitive function. The data in this paper comes from Taipei City Hospital, Renai Branch, and National Taiwan University Hospital, including dementia patients and normal people. We use the subject's monologue description of the "cookie theft" diagram. We expect that the lack of some brain functions of dementia patients will affect the language expression ability and speech fluency. This research will analyze from basic machine learning classification and then deep learning model classification. The analysis process is divided into two parts. The first part is to extract features through linguistics. We hope to classify the data by part of speech distribution and related features. In the second part, we add contextual considerations through word embedding. And introduce techniques like BERT, to find an algorithm that can effectively detect patients with dementia.

並列關鍵字

Dementia ； Feature engineering ； Natural Language Processing ； Machine Learning ； Deep Learning

參考文獻

1. Luz, S., de la Fuente, S., Albert, P. (2018). A method for analysis of patient speech in dialogue for dementia detection. arXiv preprint arXiv:1811.09919.

Google Scholar

2. Ieracitano, C., Mammone, N., Hussain, A., Morabito, F. C. (2020). A novel multi-modal machine learning based approach for automatic classification of EEG recordings in dementia. Neural Networks, 123, 176-190.

Google Scholar

3. Tanaka, H., Adachi, H., Kazui, H., Ikeda, M., Kudo, T., Nakamura, S. (2019, October). Detecting Dementia from Face in Human-Agent Interaction. In Adjunct of the 2019 International Conference on Multimodal Interaction (pp. 1-6).

Google Scholar

4. 陳奕翔. (2020). 失智長者之語音資料庫建立與應用. 臺灣大學電信工程學研究所學位論文, 1-48.

Google Scholar

5. Campbell, E. L., Docío-Fernández, L., Raboso, J. J., García-Mateo, C. (2020). Alzheimer's Dementia Detection from Audio and Text Modalities. arXiv preprint arXiv:2008.04617.

Google Scholar

國際替代計量

基於語料的失智患者辨識系統

未授權

主題瀏覽