Computational Diagnosis of Parkinson's Disease Directly from Natural Speech Using Machine Learning Techniques

@article{Frid2014ComputationalDO,
  title={Computational Diagnosis of Parkinson's Disease Directly from Natural Speech Using Machine Learning Techniques},
  author={Alex Frid and Hananel Hazan and Dan Hilu and L. Manevitz and Lorraine O. Ramig and Shimon Sapir},
  journal={2014 IEEE International Conference on Software Science, Technology and Engineering},
  year={2014},
  pages={50-53}
}
  • A. Frid, Hananel Hazan, S. Sapir
  • Published 11 June 2014
  • Computer Science
  • 2014 IEEE International Conference on Software Science, Technology and Engineering
The human voice signal carries much information in addition to direct linguistic semantic information. This information can be perceived by computational systems. In this work, we show that early diagnosis of Parkinson's disease is possible solely from the voice signal. This is in contrast to earlier work in which we showed that this can be done using hand-calculated features of the speech (such as formants) as annotated by professional speech therapists. In this paper, we review that work and… 

Figures and Tables from this paper

Diagnosis of Parkinson's disease from continuous speech using deep convolutional networks without manual selection of features
TLDR
This work focuses on automating the process of diagnosis from continuous native speech by removing the necessity of a trained personal from the diagnosis process by using an adaptation of Convolutional Neural Network architecture for one-dimensional signal processing on a relatively small training set.
Estimating the severity of parkinson's disease from speech using linear regression and database partitioning
TLDR
The paper presents the method of linear regression models on a set of extracted acoustic features from the middle of vowels in words, sentences and continuous speech, and the partitioning of the speech samples according to their total length into parts with long, medium and short duration.
Automatic estimation of severity of Parkinson's disease based on speech rhythm related features
TLDR
A method for Parkinson's disease severity level estimation based on speech rhythm related features extracted from running speech uttered by Hungarian Parkinson patients and healthy control population is proposed.
Evaluation of Machine learning and Deep learning algorithms combined with dimentionality reduction techniques for classification of Parkinson’s Disease
TLDR
A versatile range of classification based machine learning and deep learning algorithms, employed with different dimensionality reduction (DR) techniques are used and a comparative study of their accuracies in differentiating a healthy human from one who is afflicted by the disease is performed.
Parkinson’s Disease Detection using Machine Learning Techniques
TLDR
The experimental results of PD based detection system prove that the proposed detection method achieves a higher accuracy of the diagnosis of PD patients from healthy people.
The Diagnosis of Parkinson's Disease Based on Gait, Speech Analysis and Machine Learning Techniques
TLDR
This review presents that machine learning has good performances for the diagnosis of PD, however, it can only be a diagnosis tool to help doctors because of its limited generalization.
Parkinson’s Disease Severity Estimation on Hungarian Speech Using Various Speech Tasks
TLDR
The investigated features seem to be useful and relevant for the automatic diagnosis of Parkinson’s disease based on the classification and regression performances.
Parkinson’s Disease Identification using Glottal flow analysis
TLDR
SEDREAMS is used to diagnose Parkinson’s disease early with precise detection of Glottal closing and opening instants and showed an accuracy of 88.7%.
Classification of Parkinson’s disease utilizing multi-edit nearest-neighbor and ensemble learning algorithms with speech samples
TLDR
A proposed PD classification algorithm that combines a multi-edit-nearest-neighbor (MENN) algorithm and an ensemble learning algorithm that can be applied to future studies seeking to improve PD classification methods is shown.
Multi-Class Classification in Parkinson’s Disease by Leveraging Internal Topological Structure of the Data and of the Label Space
TLDR
On the Parkinson classification task, these methods obtained results which are comparable to the best accuracy on simple two class classification, which seems relatively generic and should be applicable to many data sets.
...
...

References

SHOWING 1-10 OF 29 REFERENCES
Early diagnosis of Parkinson's disease via machine learning on speech data
TLDR
It is shown that machine learning tools can be used for the early diagnosis of Parkinson's disease from speech data and that while the training phase of machine learning process from one country can be reused in the other; different features dominate in each country; presumably because of languages differences.
Quantitative acoustic measurements for characterization of speech and voice disorders in early untreated Parkinson's disease.
TLDR
It was found that measurement of the fundamental frequency variations applied to two selected tasks was the best method for separating healthy from PD subjects and it has been demonstrated that 78% of early untreated PD subjects indicate some form of vocal impairment.
Vowel articulation in Parkinson's disease.
Speech and swallowing disorders in Parkinson disease
TLDR
A rat model for studying neuropharmacologic effects on vocalization in Parkinson disease has been developed and changes in brain activity due to LSVT LOUD provide preliminary evidence for neural plasticity.
Acoustic-phonetic analysis of fricatives for classification using SVM based algorithm
  • A. Frid, Yizhar Lavner
  • Computer Science
    2010 IEEE 26-th Convention of Electrical and Electronics Engineers in Israel
  • 2010
TLDR
An effective algorithm for classification of one group of phonemes, namely the unvoiced fricatives, which are characterized by a relatively large amount of spectral energy in the high frequency range is presented.
Intensive voice treatment in Parkinson’s disease: Lee Silverman Voice Treatment
TLDR
The role of speech therapy in the landscape of exercise-based interventions for individuals with PD is examined, focusing on the intensive voice treatment protocol, Lee Silverman Voice Treatment, as an example therapy.
Formant centralization ratio: a proposal for a new acoustic measure of dysarthric speech.
TLDR
The present findings indicate that the FCR is a sensitive, valid, and reliable acoustic metric for distinguishing dysarthric from unimpaired speech and for monitoring treatment effects, probably because of reduced sensitivity to interspeaker variability and enhanced sensitivity to vowel centralization.
On the use of autocorrelation analysis for pitch detection
TLDR
Several types of (nonlinear) preprocessing which can be used to effectively spectrally flatten the speech signal are presented and an algorithm for adaptively choosing a frame size for an autocorrelation pitch analysis is discussed.
Acoustic-phonetic features for the automatic classification of fricatives.
TLDR
A statistically guided, knowledge-based, acoustic-phonetic system for the automatic classification of fricatives in speaker-independent continuous speech is proposed, which uses an auditory-based front-end processing system and incorporates new algorithms for the extraction and manipulation of the acoustic- phonetic features that proved to be rich in their information content.
Diagnosis and treatment of Parkinson's disease: a systematic review of the literature.
TLDR
MetaWorks investigators developed an evidence base through a systematic review of the Englishlanguage literature from 1990 to 2000 pertinent to patients with PD, intended to serve as an information resource for decisionmakers and developers of practice guidelines and recommendations.
...
...