Parkinson’s Disease Recognition by Speech Acoustic Parameters Classification

Abstract

Thanks to improvement of means of communication performance and intelligent systems, research works to detect speech disorders by analysing voice signals are very promising. This paper demonstrates that dysarthria in people with Parkinson's disease (PWP) can be diagnosed using a classification of the characteristics of their voices. For this purpose, we have experimented two types of classifiers, namely Bernoulli and multinomial naïve Bayes in order to select the most pertinent features parameters for diagnosing PWP. The prediction accuracy achieved by using multinomial naive Bayes (NB) classifier model reaching 95 % is very encouraging.

Things to Consider When Automatically Detecting Parkinson’s Disease Using the Phonation of Sustained Vowels: Analysis of Methodological Issues

“…It is uncommon to find studies which employ different corpora in model training and testing (cross-corpora validation) [22]. In some cases, the authors do not use cross-validation, and divide the corpus into training and testing subsets randomly [14,[22][23][24][25][26]. This increases the uncertainty of the results when using small corpora (usually no more than 3 h of recordings), as the testing partition is not large enough to be considered representative.…”

Assessment of Acoustic Features and Machine Learning for Parkinson’s Detection

“…Therefore, the proposed collaborative features on Naïve Bayes is a practical approach to Parkinson's detection. At the final stage of analysis, the proposed collaborative features-based Parkinson's detection system has been compared with the current state-of-the-art function-based methods, viz., Avuçlu and Elen [ 18 ], Bourouhou et al [ 19 ], Zhang et al [ 20 ], Meghraoui et al [ 21 ], Kadiri et al [ 22 ], Polat and Nour [ 25 ], Xiong and Lu [ 26 ] and Mekyska et al [ 28 ]. Since our approach is based on a function-based approach, most of the methods taken for comparison belong to function-based approaches such as Naïve Bayes and Support Vector Machine (SVM).…”

“…Finally, an extensive discussion has been carried out regarding the shortcoming and future direction of the proposed Parkinson's detection model. [18] Naïve Bayes 22 70.26 Bourouhou et al [19] Naïve Bayes 26 65.00 Zhang et al [20] Naïve Bayes 22 69.24 Meghraoui et al [21] Bernoulli Naïve Bayes 3 62.50 Kadiri et al [22] Support Vector Machine -73.32 Polat and Nour [25] Linear Regression 45 77.50 Xiong and Lu [26] Naïve Bayes 8 72.00 Mekyska et al [28] Classification and regression trees 8 75.19 Collaborative PD (proposed) Naïve Bayes 7 78.97…”

“…With the 22 vocal features, the Naïve Bayes reveals 69.24% of detection accuracy with a 96.02% of the precision rate. Meghraoui et al [ 21 ] proposed Bernoulli and Multinomial Naïve Bayes (BMNB) on harmonicity, pitch, and pulse features. The BMNB approaches are proved to be a better solution to detect the presence of Parkinson's.…”

Acoustic and prosodic information for home monitoring of bipolar disorder

“…jitter, shimmer and harmonics-to-noise ratio) some of the most used for these purposes. 21 By way of example, jitter and shimmer have been recently used for the detection of Parkinson's disease, [22][23][24][25] Alzheimer, 26,27 post-traumatic stress, 28 multiple sclerosis and dysarthria, 29,30 thyroid patients 31 and coronary heart disease, 32 among many others.…”