Classification of speech intelligibility in Parkinson's disease

• Autorzy: Khan T. , Westin J. , Dougherty M.

Identyfikatory

DOI

10.1016/j.bbe.2013.10.003

• Języki publikacji: EN

Abstrakty

EN

A problem in the clinical assessment of running speech in Parkinson's disease (PD) is to track underlying deficits in a number of speech components including respiration, phonation, articulation and prosody, each of which disturbs the speech intelligibility. A set of 13 features, including the cepstral separation difference and Mel-frequency cepstral coefficients were computed to represent deficits in each individual speech component. These features were then used in training a support vector machine (SVM) using n-fold cross validation. The dataset used for method development and evaluation consisted of 240 running speech samples recorded from 60 PD patients and 20 healthy controls. These speech samples were clinically rated using the Unified Parkinson's Disease Rating Scale Motor Examination of Speech (UPDRS-S). The classification accuracy of SVM was 85% in 3 levels of UPDRS-S scale and 92% in 2 levels with the average area under the ROC (receiver operating characteristic) curves of around 91%. The strong classification ability of selected features and the SVM model supports suitability of this scheme to monitor speech symptoms in PD.

Słowa kluczowe

EN

Parkinson's disease; speech processing; dysarthria; support vector machine (SVM); telemonitoring

PL

choroba Parkinsona; przetwarzanie mowy; dyzartria; maszyna wektorów wspierających; telemonitoring

• Wydawca: Nałęcz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences ; Elsevier
• Czasopismo: Biocybernetics and Biomedical Engineering
• Rocznik: 2014
• Tom: Vol. 34, no. 1
• Strony: 35--45
• Opis fizyczny: Bibliogr. 34 poz., tab., wykr.

Twórcy

autor

Khan T.

• School of Innovation, Design and Technology, Malardalen University, Vasteras, Sweden; School of Technology and Business Studies, Computer Engineering, Dalarna University, Falun, Sweden

autor

Westin J.

• School of Technology and Business Studies, Computer Engineering, Dalarna University, Falun, Sweden

autor

Dougherty M.

• School of Technology and Business Studies, Computer Engineering, Dalarna University, Falun, Sweden

Bibliografia

• [1] Olanow CW, Stern MB, Sethi K. The scientific and clinical basis for the treatment of Parkinson's disease. Neurology 2009;72:s. 1–36.
• [2] Goetz CG, Stebbins GT, Wolff D, DeLeeuw W, Bronte-Stewart H, Elble R, et al. Testing objective measures of motor impairment in early Parkinson's disease: feasibility study of an at-home testing device. Mov Disord 2009;24:551–6.
• [3] Fahn S, Elton R, The UPDRS Development Committee. Unified Parkinson's disease rating scale. Recent Dev Parkinson's Dis 1987;2:153–63.
• [4] Pinto S, Ozsancak C, Tripoliti E, Thobois S, Dowsey PL, Auzou P. Treatments for dysarthria in Parkinson's disease. Lancet Neurol 2004;3(9):547–56.
• [5] Londono JDA, Llorente JIG, Lechon NS, Ruiz VO, Dominguez GC. Automatic detection of pathological voices using complexity measures, noise parameters, and Mel-cepstral coefficients. IEEE Trans Bio-Med Eng 2011;58(2):370–8.
• [6] Rusz J, Cmejla R, Ruzickova H, Ruzicka E. Acoustic analysis of voice and speech characteristics in early untreated Parkinson's disease. In: Proc. 7th Intl. Workshop on MAVEBA. Firenze University Press; 2011. p. 181–4. 77.
• [7] Gelzinis A, Verikas A, Bacauskiene M. Automated speech analysis applied to laryngeal disease categorization. Comput Methods Programs Biomed 2008;91(1):36–47.
• [8] Rusz J, Cmejla R, Ruzickova H, Ruzicka E. Objectification of dysarthria in Parkinson's disease using Bayes Theorem. In: Proc. 10th WSEAS. Vouliagmeni, Athens 2011. pp. 165–9.
• [9] Zraick R, Dennie TM, Tabbal SM, Hutton TJ, Hicks GM, Sullivan PS. Reliability of speech intelligibility ratings using the Unified Parkinson Disease Rating Scale. J Med Speech Lang Pathol 2003;11(4):227–40.
• [10] Looze CD, Ghio A, Scherer S, Pouchoulin G, Viallet F. Automatic analysis of the prosodic variations in Parkinsonian read and semi-spontaneous speech. In: Proc. Int. Conf. Speech Prosody. China: Tongji University Press; 2012.
• [11] Paja MS, Falk TH. Automated dysarthria severity classification for improved objective intelligibility assessment of spastic dysarthric speech. In: Proc. 13th Annu. Conf. ISCA. Portland, Oregon 2012.
• [12] Llorente JIG, Fraile I, Lechón RS, Ruiz NO, Vilda VGP. Automatic detection of voice impairments from textdependent running speech. Biomed Signal Process Control 2009;4(3):176–82.
• [13] Scholkopf B, Platt JC, Shawe-Taylor J, Smola AJ, Williamson RC. Estimating the support of a high-dimensional distribution. Neural Comput 2001;13(7):1443–71.
• [14] Khan T, Westin J, Dougherty M. Cepstral separation difference: a novel approach for speech impairment quantification in Parkinson's disease. Biocybern Biomed Eng 2014;34(1):25–34.
• [15] No C. International Phonetic Association, editor. Handbook of the international phonetic association: a guide to the use of the international phonetic alphabet. Cambridge: Cambridge University Press; 1999.
• [16] Silbert N, Jong KD. Focus, prosodic context, and phonological feature specification: patterns of variation in fricative production. J Acoust Soc Am 2008;5:2769–79.
• [17] Flanagan JL, Ishizaka K, Shipley KL. Synthesis of speech from a dynamic model of the vocal cords and vocal tract. Bell Syst Technol J 1975;54:485–506.
• [18] Murphy P. Source-filter comparison of measurements of fundamental frequency perturbation and amplitude perturbation for synthesized voice signals. J Voice 2008;22:125–37.
• [19] Murdoch BE, editor. Dysarthria: a physiological approach to assessment and treatment. Cheltenham, UK: Stanley Thornes; 1998.
• [20] Midi I, Dogan M, Koseoglu M, Can G, Sehitoglu MA, Gunal DI. Voice abnormalities and their relation with motor dysfunction in Parkinson's disease. Acta Neurol Scand 2008;117:26–34.
• [21] Kim S, Eriksson T, Kang HG. On the time variability of vocal tract for speaker recognition. In: 8th ICSLP. Jeju Island, Korea 2004.
• [22] Freed D. Motor speech disorders: diagnosis and treatment, 2nd ed., New York: Delmar; 2012.
• [23] Stevens S, Volkman J. The relation of pitch to frequency: a revised scale. Am J Psychol 1940;53(3):329–53.
• [24] Jones HN. Prosody in Parkinson's disease. Perspect Neurophysiol Neurogenic Speech Lang Disord 2009;19 (3):71–6.
• [25] Rosen KM, Kent RD, Delaney AL. Parametric quantitative acoustic analysis of conversation produced by speakers with dysarthria and healthy speakers. J Speech Lang Hear Res 2006;49(2):395–411.
• [26] Le PN, Ambikairajah E, Epps J, Sethu V, Choi EH. Investigation of spectral centroid features for cognitive load classification. Speech Commun 2011;53(4):540–51.
• [27] Ahmad S, Spanias AS. Cepstrum-based pitch detection using a new statistical V/UV classification algorithm. IEEE Trans Speech Audio Process 1999;7(3):333–8.
• [28] Ma EM, Yiu EL. Suitability of acoustic perturbation measures in analyzing periodic and nearly periodic voice signals. Folia Phoniatr Logop 2005;57(1):38–47.
• [29] Guttman L. A basis for scaling qualitative data. Am Sociol Rev 1944;9(2):139–50.
• [30] Berger YG. A jackknife variance estimator for uni-stage stratified samples with unequal probabilities. Biometrika 2007;94:953–64.
• [31] Guan W. New support vector machine formulations and algorithms with application to biomedical data analysis. PhD thesis. Georgia Institute of Technology; 2011.
• [32] Ustun B, Melssen WJ, Buydens LMC. Facilitating the application of support vector regression by using a universal Pearson VII function based kernel. Chemom Intell Lab Syst 2006;81(1):29–40.
• [33] Stone M. Cross-validatory choice and assessment of statistical predictions. J Roy Statist Soc Ser B (Methodological) 1974;36:111–47.
• [34] Metz CE. Basic principles of ROC analysis. Semin Nucl Med 1978;8(4):283–98.