Fully Automatic Speech-Based Analysis of the Semantic Verbal Fluency Task

Abstract

Background: Semantic verbal fluency (SVF) tests are routinely used in screening for mild cognitive impairment (MCI). In this task, participants name as many items as possible of a semantic category under a time constraint. Clinicians measure task performance manually by summing the number of correct words and errors. More fine-grained variables add valuable information to clinical assessment, but are time-consuming. Therefore, the aim of this study is to investigate whether automatic analysis of the SVF could provide these as accurate as manual and thus, support qualitative screening of neurocognitive impairment. Methods: SVF data were collected from 95 older people with MCI (n = 47), Alzheimer’s or related dementias (ADRD; n = 24), and healthy controls (HC; n = 24). All data were annotated manually and automatically with clusters and switches. The obtained metrics were validated using a classifier to distinguish HC, MCI, and ADRD. Results: Automatically extracted clusters and switches were highly correlated (r = 0.9) with manually established values, and performed as well on the classification task separating HC from persons with ADRD (area under curve [AUC] = 0.939) and MCI (AUC = 0.758). Conclusion: The results show that it is possible to automate fine-grained analyses of SVF data for the assessment of cognitive decline.

Improving the Assessment of Mild Cognitive Impairment in Advanced Age With a Novel Multi-Feature Automated Speech and Language Analysis of Verbal Fluency

“…Others have used ASR techniques to examine the VF test. In Pakhomov et al (2015), the same Kaldi ASR toolkit (Povey et al, 2011) and in König et al (2018), Google's Automatic Speech Recognition (ASR) service were used for automatic transcription of responses. These studies either attempt to predict the raw VF score based on automatically generated response (Pakhomov et al, 2015) or only investigate count-based measures beside the raw VF score for differentiating MCI from cognitively intact participants (König et al, 2018).…”

“…In Pakhomov et al (2015), the same Kaldi ASR toolkit (Povey et al, 2011) and in König et al (2018), Google's Automatic Speech Recognition (ASR) service were used for automatic transcription of responses. These studies either attempt to predict the raw VF score based on automatically generated response (Pakhomov et al, 2015) or only investigate count-based measures beside the raw VF score for differentiating MCI from cognitively intact participants (König et al, 2018). In contrast, the crux of work that differentiates it from these studies is how we employ the ASR system not only for automatic transcription but to perform the "forced alignment" algorithm for quantifying the temporal properties of verbal responses leading to the extraction of time-based measures.…”

Temporal Analysis of the Semantic Verbal Fluency Task in Persons with Subjective and Mild Cognitive Impairment

“…In the best case, the performance of the SVM with rbf kernel improved from AU C = 0.62 to AU C = 0.72 with temporal analysis. While this study was based on manually transcribed data, previous research shows that this type of analysis can be done fully automatically including ASR, which allows for easy scaling of the task (König et al, 2018).…”

An Assessment of Paralinguistic Acoustic Features for Detection of Alzheimer's Dementia in Spontaneous Speech

“…Several commonly used cognitive tests for dementia diagnosis involve linguistic assessment, These include the Mini-Mental State Examination (MMSE) [15], the five-word test [16], the frontal assessment battery [17], and the instrumental activities of daily living scale [18]. Speech continuity, for instance, may be assessed through picture description tasks [19] or through countdown tasks [20], and Semantic Verbal Fluency (SVF) usually involves naming tasks [21]. However, whilst still valuable for diagnosis, most of these neuropsychological tests offer little insight into early stages of neurodegeneration.…”