Complexity Measures of Voice Recordings as a Discriminative Tool for Parkinson’s Disease
| Autor: | Rekha Viswanathan, Peter A. Kempster, Sanjay Raghav, Adrian Bingham, Sridhar P. Arjunan, Dinesh Kumar, Beth Jelfs |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2019 |
| Předmět: |
fractal dimension
Parkinson's disease Support Vector Machine Speech recognition Clinical Biochemistry 02 engineering and technology Biosensing Techniques Article 03 medical and health sciences Dysarthria 0302 clinical medicine Discriminative model normalised mutual information dysarthria Phonetics Statistical significance medicine Humans Mathematics Aged Anova test Parkinson Disease General Medicine Mutual information 021001 nanoscience & nanotechnology medicine.disease nervous system diseases Support vector machine sustained phonemes Detrended fluctuation analysis Parkinson’s disease Voice medicine.symptom 0210 nano-technology complexity 030217 neurology & neurosurgery |
| Zdroj: | Biosensors Volume 10 Issue 1 |
| ISSN: | 2079-6374 |
| Popis: | In this paper, we have investigated the differences in the voices of Parkinson&rsquo s disease (PD) and age-matched control (CO) subjects when uttering three phonemes using two complexity measures: fractal dimension (FD) and normalised mutual information (NMI). Three sustained phonetic voice recordings, /a/, /u/ and /m/, from 22 CO (mean age = 66.91) and 24 PD (mean age = 71.83) participants were analysed. FD was first computed for PD and CO voice recordings, followed by the computation of NMI between the test groups: PD&ndash CO, PD&ndash PD and CO&ndash CO. Four features reported in the literature&mdash normalised pitch period entropy (Norm. PPE), glottal-to-noise excitation ratio (GNE), detrended fluctuation analysis (DFA) and glottal closing quotient (ClQ)&mdash were also computed for comparison with the proposed complexity measures. The statistical significance of the features was tested using a one-way ANOVA test. Support vector machine (SVM) with a linear kernel was used to classify the test groups, using a leave-one-out validation method. The results showed that PD voice recordings had lower FD compared to CO (p < 0.008). It was also observed that the average NMI between CO voice recordings was significantly lower compared with the CO&ndash PD and PD&ndash PD groups (p < 0.036) for the three phonetic sounds. The average NMI and FD demonstrated higher accuracy (> 80%) in differentiating the test groups compared with other speech feature-based classifications. This study has demonstrated that the voices of PD patients has reduced FD, and NMI between voice recordings of PD&ndash CO and PD&ndash PD is higher compared with CO&ndash CO. This suggests that the use of NMI obtained from the sample voice, when paired with known groups of CO and PD, can be used to identify PD voices. These findings could have applications for population screening. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|