Speech Based Estimation of Parkinson’s Disease Using Gaussian Processes and Automatic Relevance Determination

Despotović, Vladimir; Škovránek, Tomáš; Schommer, Christoph

doi:10.1016/j.neucom.2020.03.058

Cited by 45 publications

(21 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Voice has already proved to be a potent digital biomarker for early detection and monitoring the disease progress of various medical conditions, with the most prominent examples being the neurological disorders such as Parkinson's disease [ 6 , 7 ], Mild Cognitive Impairment and Alzheimer's disease [ 8 , 9 ], Multiple Sclerosis [ 10 ] and Amyotrophic Lateral Sclerosis [ 11 , 12 ]. Other conditions that affect voice include Rheumatoid Arthritis [ 13 , 14 ], which may lead to voice hoarseness due to cricoarytenoid joint involvement, or Diabetes Mellitus that provokes vocal fatigue caused by decreased laryngeal muscle strength in the presence of neuropathy [ 15 ].…”

Section: Introductionmentioning

confidence: 99%

Detection of COVID-19 from voice, cough and breathing patterns: Dataset and preliminary results

Despotović¹,

Ismaël²,

Cornil³

et al. 2021

Computers in Biology and Medicine

Self Cite

View full text Add to dashboard Cite

COVID-19 heavily affects breathing and voice and causes symptoms that make patients’ voices distinctive, creating recognizable audio signatures. Initial studies have already suggested the potential of using voice as a screening solution. In this article we present a dataset of voice, cough and breathing audio recordings collected from individuals infected by SARS-CoV-2 virus, as well as non-infected subjects via large scale crowdsourced campaign. We describe preliminary results for detection of COVID-19 from cough patterns using standard acoustic features sets, wavelet scattering features and deep audio embeddings extracted from low-level feature representations (VGGish and OpenL3). Our models achieve accuracy of 88.52%, sensitivity of 88.75% and specificity of 90.87%, confirming the applicability of audio signatures to identify COVID-19 symptoms. We furthermore provide an in-depth analysis of the most informative acoustic features and try to elucidate the mechanisms that alter the acoustic characteristics of coughs of people with COVID-19.

show abstract

Section: Introductionmentioning

confidence: 99%

Detection of COVID-19 from voice, cough and breathing patterns: Dataset and preliminary results

Despotović¹,

Ismaël²,

Cornil³

et al. 2021

Computers in Biology and Medicine

Self Cite

View full text Add to dashboard Cite

show abstract

“…Authors claimed that reducing dimensionality plays an important role in improving overall classification of PD. Authors in [7] introduced the combination of Gaussian processes and automatic relevance determination for detecting PD. The study was conducted on two PD dataset and the focus was based on the using small amount of relevant acoustic features for detection.…”

Section: A Related Studies On Speech Impairmentmentioning

confidence: 99%

“…It is majorly caused by the lack of dopamine (neurotransmitter) in the human brain [4] and its effect can be categorized into motor and non-motor symptoms such as voice/speech impairment, dementia, depression, slow thinking, rigidity, tremor, bradykinesia, and other cognitive disabilities [4][5]. From 60% to 90% of PD P patients suffer from speech impairment such as slurred, mumbled or slow speech [6][7], among other symptoms.…”

Section: Introductionmentioning

confidence: 99%

BiLSTM with Data Augmentation using Interpolation Methods to Improve Early Detection of Parkinson Disease

Abayomi‐Alli

Damaševičius

Maskeliūnas

et al. 2020

Proceedings of the 2020 Federated Conference on Computer Science and Information Systems

View full text Add to dashboard Cite

The lack of dopamine in the human brain is the cause of Parkinson disease (PD) which is a degenerative disorder common globally to older citizens. However, late detection of this disease before the first clinical diagnosis has led to increased mortality rate. Research effort towards the early detection of PD has encountered challenges such as: small dataset size, class imbalance, overfitting, high false detection rate, model complexity, etc. This paper aims to improve early detection of PD using machine learning through data augmentation for very small datasets. We propose using Spline interpolation and Piecewise Cubic Hermite Interpolating Polynomial (Pchip) interpolation methods to generate synthetic data instances. We further investigate on reducing dimensionality of features for effective and real-time classification while considering computational complexity of implementation on real-life mobile phones. For classification we use Bidirectional LSTM (BiLSTM) deep learning network and compare the results with traditional machine learning algorithms like Support Vector Machine (SVM), Decision Tree, Logistic regression, KNN and Ensemble bagged tree. For experimental validation we use the Oxford Parkinson disease dataset with 195 data samples, which we have augmented with 571 synthetic data samples. The results for BiLSTM shows that even with a holdout of 90%, the model was still able to effectively recognize PD with an average accuracy for ten rounds experiment using 22 features as 82.86%, 97.1%, and 96.37% for original, augmented (Spline) and augmented (Pchip) datasets, respectively. Our results show that proposed data augmentation schemes have significantly (p < 0.001) improved the accuracy of PD recognition on a small dataset using both classical machine learning models and BiLSTM.

show abstract

“…However, UPDRS has a disadvantage that it takes a long time to measure for the treatment and observation of PD patients Despotovic et al proposed a model to detect the PD using Gaussian processes, automatic relevance determination, and speech impairment, which is a symptom of PD. 20 Solana-Lavalle et al also proposed vocal-based PD detection in the early stages of PD disease, using support vector machine (SVM), k-nearest neighbor (KNN), multi-layer perceptron (MLP), and random forest (RF). 21 Diaz et al used dynamically enhanced static handwriting for building computer-aided PD diagnosis systems.…”

Section: Related Workmentioning

confidence: 99%

Neuro-Fuzzy System for Detecting Pd Patients Based on Euclid Distance, Fft, and Pca

Jang

Lee

2020

J. Mech. Med. Biol.

View full text Add to dashboard Cite

This study proposes a method to distinguish between healthy people and Parkinson’s disease patients using sole pressure sensor data, neural network with weighted fuzzy membership (NEWFM), and preprocessing techniques. The preprocessing techniques include fast Fourier transform (FFT), Euclidean distance, and principal component analysis (PCA), to remove noise in the data for performance enhancement. To make the features usable as inputs for NEWFM, the Euclidean distances between the left and right sole pressure sensor data were used at the first step. In the second step, the frequency scales of the Euclidean distances extracted in the first step were divided into individual scales by the FFT using the Hamming method. In the final step, 1–15 dimensions were extracted as the features of NEWFM from the individual scales by the FFT extracted in the second step by the PCA. An accuracy of 75.90% was acquired from the eight dimensions as the inputs of NEWFM.

show abstract

Speech Based Estimation of Parkinson’s Disease Using Gaussian Processes and Automatic Relevance Determination

Cited by 45 publications

References 19 publications

Detection of COVID-19 from voice, cough and breathing patterns: Dataset and preliminary results

Detection of COVID-19 from voice, cough and breathing patterns: Dataset and preliminary results

BiLSTM with Data Augmentation using Interpolation Methods to Improve Early Detection of Parkinson Disease

Neuro-Fuzzy System for Detecting Pd Patients Based on Euclid Distance, Fft, and Pca

Contact Info

Product

Resources

About