König Alexandra, Satt Aharon, Sorin Alexander, Hoory Ron, Toledo-Ronen Orith, Derreumaux Alexandre, Manera Valeria, Verhey Frans, Aalten Pauline, Robert Phillipe H, David Renaud
Research Unit CoBTeK - Cognition Behaviour Technology, Edmond & Lily Safra Research Center, University of Nice Sophia Antipolis, Nice, France; Alzheimer Centre Limburg, Maastricht University Medical Center, School for Mental Health and Neuroscience, Maastricht, The Netherlands.
Speech Technologies, IBM Research, Haifa, Israel.
Alzheimers Dement (Amst). 2015 Mar 29;1(1):112-24. doi: 10.1016/j.dadm.2014.11.012. eCollection 2015 Mar.
To evaluate the interest of using automatic speech analyses for the assessment of mild cognitive impairment (MCI) and early-stage Alzheimer's disease (AD).
Healthy elderly control (HC) subjects and patients with MCI or AD were recorded while performing several short cognitive vocal tasks. The voice recordings were processed, and the first vocal markers were extracted using speech signal processing techniques. Second, the vocal markers were tested to assess their "power" to distinguish among HC, MCI, and AD. The second step included training automatic classifiers for detecting MCI and AD, using machine learning methods and testing the detection accuracy.
The classification accuracy of automatic audio analyses were as follows: between HCs and those with MCI, 79% ± 5%; between HCs and those with AD, 87% ± 3%; and between those with MCI and those with AD, 80% ± 5%, demonstrating its assessment utility.
Automatic speech analyses could be an additional objective assessment tool for elderly with cognitive decline.
评估使用自动语音分析来评估轻度认知障碍（MCI）和早期阿尔茨海默病（AD）的价值。
在健康老年对照（HC）受试者以及MCI或AD患者执行多项简短认知语音任务时进行录音。对语音记录进行处理，并使用语音信号处理技术提取首批语音标志物。其次，对语音标志物进行测试，以评估其区分HC、MCI和AD的“能力”。第二步包括使用机器学习方法训练用于检测MCI和AD的自动分类器，并测试检测准确性。
自动音频分析的分类准确率如下：HC与MCI患者之间为79%±5%；HC与AD患者之间为87%±3%；MCI患者与AD患者之间为80%±5%，证明了其评估效用。
自动语音分析可能是认知功能减退老年人的一种额外客观评估工具。
