Learning diagnostic models using speech and language measures

This pilot study evaluates the ability of machined learned algorithms to assist with the differential diagnosis of dementia subtypes based on brief (< 10 min) spontaneous speech samples. We analyzed 1 recordings of a brief spontaneous speech sample from 48 participants from 5 different groups: 4 types of dementia plus healthy controls. Recordings were analyzed using a speech recognition system optimized for speakerindependent spontaneous speech. Lexical and acoustic features were automatically extracted. The resulting feature profiles were used as input to a machine learning system that was trained to identify the diagnosis assigned to each research participant. Between groups lexical and acoustic differences features were detected in accordance with expectations from prior research literature suggesting that classifications were based on features consistent with human-observed symptomatology. Machine learning algorithms were able to identify participants' diagnostic group with accuracy comparable to existing diagnostic methods in use today. Results suggest this clinical speech analytic approach offers promise as an additional, objective and easily obtained source of diagnostic information for clinicians.

show abstract

“…See Table 4, which shows the performance on distinguishing FTLD subtypes. For more detail on machine learning results see Peintner et al (2008).…”

Section: Machine Learning Resultsmentioning

confidence: 99%

Aided diagnosis of dementia type through computer-based analysis of spontaneous speech

Jarrold¹,

Peintner²,

Wilkins³

et al. 2014

Proceedings of the Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal to Clinical Reality

Self Cite

132

124

View full text Add to dashboard Cite

show abstract

“…For example, voice has been used to reveal patterns in the voice of patients with depression [41] and speech alterations in neurological disorders [19]. Video has been used to quantify the tremor in patients with Parkinson [44].…”

Section: Extracting Information From the Health Social Webmentioning

confidence: 99%

Review of Extracting Information From the Social Web for Health Personalization

Fernández-Luque¹,

Karlsen²,

Bonander³

2011

J Med Internet Res

View full text Add to dashboard Cite

In recent years the Web has come into its own as a social platform where health consumers are actively creating and consuming Web content. Moreover, as the Web matures, consumers are gaining access to personalized applications adapted to their health needs and interests. The creation of personalized Web applications relies on extracted information about the users and the content to personalize. The Social Web itself provides many sources of information that can be used to extract information for personalization apart from traditional Web forms and questionnaires.This paper provides a review of different approaches for extracting information from the Social Web for health personalization. We reviewed research literature across different fields addressing the disclosure of health information in the Social Web, techniques to extract that information, and examples of personalized health applications. In addition, the paper includes a discussion of technical and socioethical challenges related to the extraction of information for health personalization.

show abstract

“…Various acoustic features and text features are designed and evaluated for PWA speech assessment. Peintner et al [6] performed an automatic classification on three sub-types of frontotemporal lobar degeneration in a relatively small dataset which includes the progressive non-fluent aphasia. They proposed a number of phone duration features, part-of-speech features as well as linguistic inquiry and word count features, which were extracted from ASR outputs.…”

Section: Introductionmentioning

confidence: 99%

An End-to-End Approach to Automatic Speech Assessment for Cantonese-speaking People with Aphasia

Qin

Lee

et al. 2020

J Sign Process Syst

View full text Add to dashboard Cite

Conventional automatic assessment of pathological speech usually follows two main steps: (1) extraction of pathology-specific features; (2) classification or regression on extracted features. Given the great variety of speech and language disorders, feature design is never a straightforward task, and yet it is most crucial to the performance of assessment. This paper presents an end-to-end approach to automatic speech assessment for Cantonese-speaking People With Aphasia (PWA). The assessment is formulated as a binary classification task to discriminate PWA with high scores of subjective assessment from those with low scores. The sequence-to-one Recurrent Neural Network with Gated Recurrent Unit (GRU-RNN) and Convolutional Neural Network (CNN) models are applied to realize the end-to-end mapping from fundamental speech features to the classification result. The pathology-specific features used for assessment can be learned implicitly by the neural network model. Class Activation Mapping (CAM) method is utilized to visualize how those features contribute to the assessment result. Our experimental results 2 Ying Qin et al.show that the end-to-end approach outperforms the conventional two-step approach in the classification task, and confirm that the CNN model is able to learn impairment-related features that are similar to human-designed features. The experimental results also suggest that CNN model performs better than sequence-to-one GRU-RNN model in this specific task.

show abstract

Learning diagnostic models using speech and language measures

Cited by 32 publications

References 10 publications

Aided diagnosis of dementia type through computer-based analysis of spontaneous speech

Aided diagnosis of dementia type through computer-based analysis of spontaneous speech

Review of Extracting Information From the Social Web for Health Personalization

An End-to-End Approach to Automatic Speech Assessment for Cantonese-speaking People with Aphasia

Contact Info

Product

Resources

About