Automatic remote screening of speech features associated with Alzheimer's disease

评估阿尔茨海默病（AD）是一个昂贵而费力的过程，考虑到我们迅速老龄化的人口，这是不可持续的，特别是如果考虑到间接成本，如在旅行中损失的时间、等待时间和花费在评估上的时间，这是一个非常费力的过程，通常每隔几年才重复一次。此外，阿尔茨海默病症状的高度变异性无法通过单一评估准确确定，这降低了当前实践的有效性。快速、可重复和远程评估是必不可少的。**许多AD患者的语言能力下降。语言学指标，比如代词与名词的比例，与认知能力下降密切相关。通过测量这些特征**并使用先进的机器学习软件，我们开发了计算方法**，从简短的语音样本中识别阿尔茨海默氏症的准确率超过85%。随着**我们不断改进这项技术，我们将通过公共电话**网络提供1-800号码。该系统将自动接收门诊病人的电话（并给他们打电话），并完全自动化基于语音的测试，包括问答、叙述和随机项目生成。这将被封装在一个界面**中，卫生保健工作者可以使用它来分配和监测评估和演习。**目前，要提取语言度量，语音必须由专业人员转录，这**在大规模上是不可能的。不幸的是，目前还没有针对AD患者的自动语音识别系统**。因此，我们将为AD患者**提供第一个完整的ASR系统。**我们的目标是：1)提高AD患者ASR的准确性；2)完善和**验证我们的远程监控/评估软件；3)将我们的软件嵌入到用户友好的基于手机的界面中。

Assessing Alzheimer's disease (AD) is an expensive and laborious process that is**unsustainable given our rapidly aging population, especially if one accounts for indirect costs**such as lost time in travel, wait times, and hours spent in assessment, which is a process so**labourious that it is often repeated only every few years. Moreover, the high variability of**symptoms in AD cannot be ascertained accurately from a single assessment, which**minimizes the effectiveness of current practice. Rapid, repeatable, and remote assessment is**essential.**Many individuals with AD have diminished language abilities. Linguistic measures, like the**ratio of pronouns to nouns, relate strongly to cognitive decline. By measuring such features**and using advanced machine learning software, we have developed computational methods**that are over 85% accurate in identifying Alzheimer's disease from short speech samples. As**we continue to improve this technology, we will make it available over the public phone**network accessible by a 1-800 number. This system will automatically receive calls from (and**make calls to) outpatients, and completely automate speech-based tests including questionanswering,**narratives, and random item generation. This will be encapsulated in an interface**that healthcare workers can use to assign and monitor assessments and exercises.**Currently, to extract language measures, speech must be transcribed by professionals, which**is impossible on large scales. Unfortunately, no automatic speech recognition (ASR) system**exists for people with AD. We will therefore provide the first complete ASR system for people**with AD.**Our objectives are: 1) to improve the accuracy of ASR for people with AD; 2) to refine and**validate our software in remote monitoring/assessment; 3) to embed our software into a userfriendly**phone-based interface.