Identifying Cognitive Impairment Using Sentence Representation Vectors
Identifying Cognitive Impairment Using Sentence Representation Vectors
复制标题
DOI：
10.21437/interspeech.2021-915
复制
发表时间：
2021-08
期刊：
影响因子：
--
通讯作者：
B. Mirheidari;Yilin Pan;Daniel Blackburn;R. O'Malley;H. Christensen
中科院分区：
文献类型：
--
作者：
B. Mirheidari;Yilin Pan;Daniel Blackburn;R. O'Malley;H. Christensen
广泛使用的词向量可以在句子级别进行扩展，以执行广泛的自然语言处理（NLP）任务。最近，双向编码器表示从变压器（BERT）语言表示实现了这些应用程序的最先进的性能。该模型使用标点和格式良好的（书面）文本进行训练，然而，当输入文本是自动语音识别（ASR）的错误和未标点输出时，模型的性能会显着下降。我们使用滑动窗口和平均方法对BERT的文本进行预处理，以提取用于分类与认知障碍相关的三种诊断类别的特征：神经退行性疾病（ND），轻度认知障碍（MCI）和健康对照（HC）。内部数据集包含智能虚拟代理（IVA）的音频记录，该智能虚拟代理除了给出图片描述提示外，还向参与者询问几个会话问题提示。对于三向分类，我们使用预训练的，未分类的基础BERT和双向分类器（HCvs）实现了73.88%的F分数（准确率：76.53%）。ND）的准确率为89.80%（准确率为90%）。我们进一步提高这些使用提示选择技术，达到F-分数分别为79.98%（准确率：81.63%）和93.56%（准确率：93.75%）。
The widely used word vectors can be extended at the sentence level to perform a wide range of natural language processing (NLP) tasks. Recently the Bidirectional Encoder Rep-resentations from Transformers (BERT) language representation achieved state-of-the-art performance for these applications. The model is trained with punctuated and well-formed (writ-ten) text, however, the performance of the model drops significantly when the input text is the – erroneous and un-punctuated– output of automatic speech recognition (ASR). We use a sliding window and averaging approach for pre-processing text for BERT to extract features for classifying three diagnostic categories relating to cognitive impairment: neurodegenerative dis-order (ND), mild cognitive impairment (MCI), and healthy controls (HC). The in-house dataset contains the audio recordings of an intelligent virtual agent (IVA) who asks the participants several conversational questions prompts in addition to giving a picture description prompt. For the three-way classi-fication, we achieve a 73.88% F-score (accuracy: 76.53%) us-ing the pre-trained, uncased base BERT and for the two-way classifier (HCvs. ND) we achieve 89.80% (accuracy: 90%). We further improve these by using a prompt selection technique, reaching the F-scores of 79.98% (accuracy: 81.63%) and 93.56% (accuracy:93.75%) respectively.