Abstract
In recent years, there is an increasing interest in building e-health systems. The systems built to deliver the health services with the use of internet and communication technologies aim to reduce the costs arising from outpatient visits of patients. Some of the related recent studies propose machine learning–based telediagnosis and telemonitoring systems for Parkinson’s disease (PD). Motivated from the studies showing the potential of speech disorders in PD telemonitoring systems, in this study, we aim to estimate the severity of PD from voice recordings of the patients using motor Unified Parkinson’s Disease Rating Scale (UPDRS) as the evaluation metric. For this purpose, we apply various speech processing algorithms to the voice signals of the patients and then use these features as input to a two-stage estimation model. The first step is to apply a wrapper-based feature selection algorithm, called Boruta, and select the most informative speech features. The second step is to feed the selected set of features to a decision tree–based boosting algorithm, extreme gradient boosting, which has been recently applied successfully in many machine learning tasks due to its generalization ability and speed. The feature selection analysis showed that the vibration pattern of the vocal fold is an important indicator of PD severity. Besides, we also investigate the effectiveness of using age and years passed since diagnosis as covariates together with speech features. The lowest mean absolute error with 3.87 was obtained by combining these covariates and speech features with prediction level fusion.
Similar content being viewed by others
References
Abdulhay E, Arunkumar N, Narasimhan K, Vellaiappan E, Venkatraman V (2018) Gait and tremor investigation using machine learning techniques for the diagnosis of Parkinson disease. Fut Gener Comput Syst 83:366–373
Afonso LC, Rosa GH, Pereira CR, Weber SA, Hook C, Albuquerque VHC, Papa JP (2019) A recurrence plot-based approach for Parkinson’s disease identification. Fut Gener Comput Syst 94:282–292
Al Mamun KA, Alhussein M, Sailunaz K, Islam MS (2017) Cloud based framework for Parkinson’s disease diagnosis and monitoring system for remote healthcare applications. Fut Gener Comput Syst 66:36–47
Asuncion A, Newman D (2007) UCI machine learning repository
Bayestehtashk A, Asgari M, Shafran I, McNames J (2015) Fully automated assessment of the severity of Parkinson’s disease from speech. Comput Speech Lang 29(1):172–185
Bergstra J, Bengio Y (2012) Random search for hyper-parameter optimization. J Mach Learn Res 13(Feb):281–305
Boersma P (2006) Praat: doing phonetics by computer. http://www.praat.org/
Buza K, Ágnes Varga N (2016) ParkinsoNET: Estimation of UPDRS score using hubness-aware feedforward neural networks. Appl Artif Intell 30(6):541–555
Chan P, Holford N (2001) Drug treatment effects on disease progression. Ann Rev Pharmacol Toxicol 41(1):625–659
Chen T, Guestrin C (2016) Xgboost: a scalable tree boosting system. In: Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining. ACM, pp 785–794
Chen X, Huang L, Xie D, Zhao Q (2018) EGBMMDA: extreme gradient boosting machine for miRNA-disease association prediction. Cell Death Disease 9(1):3
Chiba T, Kajiyama M (1958) The vowel: its nature and structure, vol 652. Phonetic society of Japan Tokyo
Dorsey E, Constantinescu R, Thompson J, Biglan K, Holloway R, Kieburtz K, Marshall F, Ravina B, Schifitto G, Siderowf A et al (2007) Projected number of people with Parkinson disease in the most populous nations, 2005 through 2030. Neurology 68(5):384–386
Duffy J (2013) Motor speech disorders: substrates, differential diagnosis, and management, 3 edn. Elsevier Mosby, St Louis
Espay AJ, Bonato P, Nahab FB, Maetzler W, Dean JM, Klucken J, Eskofier BM, Merola A, Horak F, Lang AE, Reilmann R, Giuffrida J, Nieuwboer A, Horne M, Little MA, Litvan I, Simuni T, Dorsey ER, Burack MA, Kubota K, Kamondi A, Godinho C, Daneault J, Mitsi G, Krinke L, Hausdorff JM, Bloem BR, Papapetropoulos S (2016) Technology in Parkinson’s disease: challenges and opportunities. Mov Disord 31(9):1272–1282
Fahn S, Elton R et al (1987) Unified parkinson’s disease rating scale. Recent Dev Parkinson’s Disease 2:153–164
Friedman J, Hastie T, Tibshirani R (2001) The elements of statistical learning, vol 1. Springer series in statistics, New York
Friedman JH (2001) Greedy function approximation: a gradient boosting machine. Annals of statistics:1189–1232
Friedman JH (2002) Stochastic gradient boosting. Comput Stat Data Anal 38(4):367–378
Gelzinis A, Verikas A, Bacauskiene M (2008) Automated speech analysis applied to laryngeal disease categorization. Comput Methods Prog Biomed 91(1):36–47
Godino-Llorente JI, Gomez-Vilda P, Blanco-Velasco M (2006) Dimensionality reduction of a pathological voice quality assessment system based on Gaussian mixture models and short-term cepstral parameters. IEEE Trans Biomed Eng 53(10):1943–1953
Goetz CG, Stebbins GT (2004) Assuring interrater reliability for the UPDRS motor section: utility of the UPDRS teaching tape. Movement Disord 19(12):1453–1456
Goetz CG, Stebbins GT, Wolff D, DeLeeuw W, Bronte-Stewart H, Elble R, Hallett M, Nutt J, Ramig L, Sanger T et al (2009) Testing objective measures of motor impairment in early Parkinson’s disease: feasibility study of an at-home testing device. Mov Disord 24(4):551–556
Guyon I, Elisseeff A (2003) An introduction to variable and feature selection. J Mach Learn Res 3:1157–1182
Harel B, Cannizzaro M, Snyder PJ (2004) Variability in fundamental frequency during speech in prodromal and incipient Parkinson’s disease: a longitudinal case study. Brain Cogn 56(1):24–29
Hartelius L, Svensson P (1994) Speech and swallowing symptoms associated with Parkinson’s disease and multiple sclerosis: a survey. Folia Phoniatrica Logopaedica 46(1):9–17
Herd CP, Tomlinson CL, Deane KH, Brady MC, Smith CH, Sackley CM, Clarke CE (2012) Speech and language therapy versus placebo or no intervention for speech problems in Parkinson’s disease. The Cochrane Library
Hindle JV (2010) Ageing, neurodegeneration and Parkinson’s disease. Age Ageing 39(2):156–161
Ho AK, Iansek R, Marigliani C, Bradshaw JL, Gates S (1999) Speech impairment in a large sample of patients with Parkinson’s disease. Behav Neurol 11(3):131–137
Jacobson BH, Johnson A, Grywalski C, Silbergleit A, Jacobson G, Benninger MS, Newman CW (1997) The voice handicap index (VHI): development and validation. Amer J Speech-Lang Pathol 6 (3):66–70
Jankovic J (2008) Parkinson’s disease: clinical features and diagnosis. Journal of Neurology. Neurosur Psych 79(4):368–376
Kowal SL, Dall TM, Chakrabarti R, Storm MV, Jain A (2013) The current and projected economic burden of Parkinson’s disease in the United States. Mov Disord 28(3):311–318
Kumar M, Pachori RB, Acharya UR (2017) Characterization of coronary artery disease using flexible analytic wavelet transform applied on ECG signals. Biomed Signal Process Control 31:301–308
Kursa M, Jankowski A, Rudnicki W (2010) Boruta – a system for feature selection. Fund Inf 101:271–285
Kursa M, Rudnicki W (2010) Feature selection with the boruta package. J Stat Softw 36 (11):1–13
Lipsmeier F, Taylor KI, Kilchenmann T, Wolf D, Scotland A, Schjodt-Eriksen J, Cheng WY, Fernandez-Garcia I, Siebourg-Polster J, Jin L et al (2018) Evaluation of smartphone-based testing to generate exploratory outcome measures in a phase 1 Parkinson’s disease clinical trial. Mov Disord 33 (8):1287–1297
Logemann JA, Fisher HB, Boshes B, Blonsky ER (1978) Frequency and cooccurrence of vocal tract dysfunctions in the speech of a large sample of Parkinson patients. J Speech Hear Disord 43(1):47–57
Magdalinou N, Morris HR (2017) Clinical features and differential diagnosis of parkinson’s disease. In: Movement disorders curricula. Springer, pp 103–115
Majdinasab F, Karkheiran S, Moradi N, Shahidi GA, Salehi M (2012) Relation between Voice Handicap Index (VHI) and disease severity in Iranian patients with Parkinson’s disease. Med J Islamic Republ Iran 26(4):157
Markaki M, Stylianou Y, Arias-Londoño JD, Godino-Llorente JI (2010) Dysphonia detection based on modulation spectral features and cepstral coefficients. In: 2010 IEEE international conference on Acoustics speech and signal processing (ICASSP). IEEE, pp 5162–5165
Mekyska J, Janousova E, Gomez-Vilda P, Smekal Z, Rektorova I, Eliasova I, Kostalova M, Mrackova M, Alonso-Hernandez JB, Faundez-Zanuy M et al (2015) Robust and complex approach of pathological speech signal analysis. Neurocomputing 167:94–111
Mekyska J, Rektorova I, Smekal Z (2011) Selection of optimal parameters for automatic analysis of speech disorders in Parkinson’s disease. In: 2011 34th international conference on Telecommunications and signal processing (TSP). IEEE, pp 408–412
Michaelis D, Gramss T, Strube HW (1997) Glottal-to-noise excitation ratio–a new measure for describing pathological voices. Acta Acust United Acust 83(4):700–706
Midi I, Dogan M, Koseoglu M, Can G, Sehitoglu MA, Gunal DI (2008) Voice abnormalities and their relation with motor dysfunction in Parkinson’s disease. Acta Neurol Scand 117(1):26–34
Miller A (2002) Subset selection in regression. CRC Press, Boca Raton
Murty KSR, Yegnanarayana B (2008) Epoch extraction from speech signals. IEEE Trans Audio Speech Lang Process 16(8):1602–1613
Naranjo L, Pérez CJ, Martín J (2017) Addressing voice recording replications for tracking Parkinson’s disease progression. Med Biol Eng Comput 55(3):365–373
Ozkanca Y, Goksu Ozturk M, Ekmekci MN, Atkins DC, Demiroglu C, Hosseini Ghomi R (2019) Depression screening from voice samples of patients affected by parkinson’s disease. Digit Biomarkers 3(2):72–82
Martínez-Martín P , Gil-Nagel A, Gracia LM, Gómez JB , Martínez-Sarriés J, Bermejo F (1994) Unified Parkinson’s disease rating scale characteristics and structure. Mov Disord 9(1):76–83
Pahwa R, Lyons KE (2013) Handbook of Parkinson’s disease. Crc Press, Boca Raton
Paja W, Wrzesień M (2013) Melanoma important features selection using random forest approach. In: 2013 6Th international conference on human system interactions (HSI). IEEE, pp 415–418
Patidar S, Pachori RB, Acharya UR (2015) Automated diagnosis of coronary artery disease using tunable-Q wavelet transform applied on heart rate signals. Knowl-Based Syst 82:1–10
Patidar S, Pachori RB, Upadhyay A, Acharya UR (2017) An integrated alcoholic index using tunable-Q wavelet transform based features extracted from EEG signals for diagnosis of alcoholism. Appl Soft Comput 50:71–78
Perry TL, Ohde RN, Ashmead DH (2001) The acoustic bases for gender identification from children’s voices. J Acoust Soc Amer 109(6):2988–2998
Poona NK, Ismail R (2014) Using Boruta-selected spectroscopic wavebands for the asymptomatic detection of Fusarium circinatum stress. IEEE J Sel Top Appl Earth Observ Remote Sens 7(9):3764–3772
Poona NK, Van Niekerk A, Nadel RL, Ismail R (2016) Random forest (RF) wrappers for waveband selection and classification of hyperspectral data. Appl Spectroscopy 70(2):322–333
Post B, Merkus MP, de Bie R, de Haan RJ, Speelman JD (2005) Unified Parkinson’s disease rating scale motor examination: are ratings of nurses, residents in neurology, and movement disorders specialists interchangeable?. Movement Disord 20(12):1577–1584
Rahn DA, Chou M, Jiang JJ, Zhang Y (2007) Phonatory impairment in parkinson’s disease: evidence from nonlinear dynamic analysis and perturbation analysis. J Voice 21(1):64–71
Ramaker C, Marinus J, Stiggelbout AM, Van Hilten BJ (2002) Systematic evaluation of rating scales for impairment and disability in Parkinson’s disease. Mov Disord 17(5):867–876
Richards M, Marder K, Cote L, Mayeux R (1994) Interrater reliability of the Unified Parkinson’s Disease Rating Scale motor examination. Mov Disord 9(1):89–91
Rusz J, Cmejla R, Ruzickova H, Ruzicka E (2011) Quantitative acoustic measurements for characterization of speech and voice disorders in early untreated Parkinson’s disease. J Acoust Soc Amer 129(1):350–367
Saeb S, Lonini L, Jayaraman A, Mohr DC, Kording KP (2017) The need to approximate the use-case in clinical machine learning. GigaScience 6(5):1–9
Sakar BE, Isenkul ME, Sakar CO, Sertbas A, Gurgen F, Delil S, Apaydin H, Kursun O (2013) Collection and analysis of a Parkinson speech dataset with multiple types of sound recordings. IEEE J Biomed Health Inf 17(4):828–834
Sakar CO, Kursun O (2010) Telediagnosis of Parkinson’s disease using measurements of dysphonia. J Med Syst 34(4):591–599
Sakar CO, Serbes G, Gunduz A, Tunc HC, Nizam H, Sakar BE, Tutuncu M, Aydin T, Isenkul ME, Apaydin H (2019) A comparative analysis of speech signal processing algorithms for Parkinson’s disease classification and the use of the tunable Q-factor wavelet transform. Appl Soft Comput 74:255–263
Schoentgen J, De Guchteneere R (1995) Time series analysis of jitter. J Phon 23(1-2):189–201
Schüpbach WMM, Corvol JC, Czernecki V, Djebara MB, Golmard JL, Agid Y, Hartmann A (2010) Segmental progression of early untreated Parkinson’s disease: a novel approach to clinical rating. J Neurol Neurosurg Psych 81(1):20–25
Selesnick IW (2011) Wavelet transform with tunable Q-factor. IEEE Trans Signal Process 59 (8):3560–3575
Serbes G, Sakar BE, Gulcur HO, Aydin N (2015) An emboli detection system based on Dual Tree Complex Wavelet Transform and ensemble learning. Appl Soft Comput 37:87–94
Shue YL (2010) The voice source in speech production: data, analysis and models. Ph.D. thesis, University of California, Los Angeles
Siderowf A, McDermott M, Kieburtz K, Blindauer K, Plumb S, Shoulson I, Group PS (2002) Test–retest reliability of the unified Parkinson’s disease rating scale in patients with early Parkinson’s disease: results from a multicenter clinical trial. Movement Disord 17(4):758–763
Simpson AP (2009) Phonetic differences between male and female speech. Lang Linguist Compass 3(2):621–640
Singhi SK, Liu H (2006) Feature subset selection bias for classification learning. In: Proceedings of the 23rd international conference on Machine learning. ACM, pp 849–856
Sánchez-Ferro Á, Elshehabi M, Godinho C, Salkovic D, Hobert MA, Domingos J, Uem JM, Ferreira JJ, Maetzler W (2016) New methods for the assessment of Parkinson’s disease (2005 to 2015): A systematic review. Mov Disord 31(9):1283–1292
Touzani S, Granderson J, Fernandes S (2018) Gradient boosting machine for modeling the energy consumption of commercial buildings. Energy Build 158:1533–1543
Tsanas A (2012) Accurate telemonitoring of Parkinson’s disease symptom severity using nonlinear speech signal processing and statistical machine learning. Ph.D. thesis, Oxford Centre for Industrial and Applied Mathematics, University of Oxford, UK
Tsanas A, Little MA, McSharry PE, Ramig LO (2010) Accurate telemonitoring of Parkinson’s disease progression by noninvasive speech tests. IEEE Trans Biomed Eng 57(4):884–893
Tsanas A, Little MA, McSharry PE, Ramig LO (2011) Nonlinear speech analysis algorithms mapped to a standard metric achieve clinically useful quantification of average Parkinson’s disease symptom severity. J R Soc Interface 8(59):842–855
Tsanas A, Little MA, McSharry PE, Spielman J, Ramig LO (2012) Novel speech signal processing algorithms for high-accuracy classification of Parkinson’s disease. IEEE Trans Biomed Eng 59(5):1264–1271
Tsanas A, Ma L, McSharry P, Ramig L (2010) New nonlinear markers and insights into speech signal degradation for effective tracking of Parkinson’s disease symptom severity. In: New nonlinear markers and insights into speech signal degradation for effective tracking of parkinson’s disease symptom severity, Krakow, pp 457–460
Ulukaya S, Serbes G, Kahya YP (2017) Overcomplete discrete wavelet transform based respiratory sound discrimination with feature and decision level fusion. Biomed Signal Process Control 38:322–336
Ulukaya S, Serbes G, Kahya YP (2019) Wheeze type classification using non-dyadic wavelet transform based optimal energy ratio technique. Comput Biol Med 104:175–182
Vaiciukynas E, Verikas A, Gelzinis A, Bacauskiene M (2017) Detecting Parkinson’s disease from sustained phonation and speech signals. Plos One 12(10):e0185613
Van Den Eeden SK, Tanner CM, Bernstein AL, Fross RD, Leimpeter A, Bloch DA, Nelson LM (2003) Incidence of Parkinson’s disease: variation by age, gender, and race/ethnicity. Amer J Epidemiol 157(11):1015–1022
Vos T et al (2017) Global, regional, and national incidence, prevalence, and years lived with disability for 328 diseases and injuries for 195 countries, 1990—2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet 390(10100):1211–1259
Weismer G, Jeng JY, Laures JS, Kent RD, Kent JF (2001) Acoustic and intelligibility characteristics of sentence production in neurogenic speech disorders. Folia Phoniatrica Logopaedica 53(1):1–18
Wenning GK, Tison F, Seppi K, Sampaio C, Diem A, Yekhlef F, Ghorayeb I, Ory F, Galitzky M, Scaravilli T et al (2004) Development and validation of the unified multiple system atrophy rating scale (UMSARS). Mov Disord 19(12):1391–1402
Whitmore LS, Davis RW, McCormick RL, Gladden JM, Simmons BA, George A, Hudson CM (2016) BiocompoundML: a general biofuel property screening tool for biological molecules using Random Forest Classifiers. Energy Fuels 30(10):8410– 8418
Zarzur AP, Duarte IS, Gonċalves GdNH, Martins MAUR (2010) Laryngeal electromyography and acoustic voice analysis in Parkinson’s disease: a comparative study. Brazil J Otorhinolaryngol 76 (1):40–43
Zhan A, Mohan S, Tarolli C et al (2018) Using smartphones and machine learning to quantify parkinson disease severity: The mobile parkinson disease score. JAMA Neurology
Zhang Y, Haghani A (2015) A gradient boosting method to improve travel time prediction. Transp Res Part C: Emerg Technol 58:308–324
Zheng A, Casari A (2018) Feature engineering for machine learning: principles and techniques for data scientists. O’Reilly Media, Inc.
Funding
This research was supported by The Scientific and Technological Research Council of Turkey (TUBITAK) grant number 215E008, https://www.tubitak.gov.tr/en.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
The study has the approval of the Clinical Research Ethics Committee of Bahcesehir University.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Tunc, H.C., Sakar, C.O., Apaydin, H. et al. Estimation of Parkinson’s disease severity using speech features and extreme gradient boosting. Med Biol Eng Comput 58, 2757–2773 (2020). https://doi.org/10.1007/s11517-020-02250-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11517-020-02250-5