Automatic intelligibility assessment of dysarthric speech using glottal parameters

Objective intelligibility assessment of dysarthric speech can assist clinicians in diagnosis of speech disorders as well as in medical treatment. This study investigates the use of glottal parameters (i.e. parameters that describe the acoustical excitation of voiced speech, the glottal flow) in the automatic intelligibility assessment of dysarthric speech. Instead of directly predicting the intelligibility of dysarthric speech using a single-stage system, the proposed method utilizes a two-stage framework. In the first stage, two-class severity classification of dysarthria is performed using support vector machines (SVMs). In the second stage, intelligibility estimation of dysarthric speech is computed using a linear regression model. Two sets of glottal parameters are explored: (1) time-domain and frequency-domain parameters and (2) parameters based on principal component analysis (PCA). Acoustic parameters proposed in a similar intelligibility prediction study by Falk et al. (2012) are used as baseline features. Evaluation results show that the two-stage framework leads to improvement in the intelligibility assessment measures (correlation and root mean square error) compared to the single-stage framework. The combination of the glottal parameters sets results in better performance in the severity classification and intelligibility estimation tasks compared to the baseline features.