Detection of Amyotrophic Lateral Sclerosis (ALS) via Acoustic Analysis

Abstract

ALS is a fatal neurodegenerative disease with no cure. Experts typically measure disease progression via the ALSFRS-R score, which includes measurements of various abilities known to decline. We propose instead the use of speech analysis as a proxy for ALS progression. This technique enables 1) frequent non-invasive, inexpensive, longitudinal analysis, 2) analysis of data recorded in the wild, and 3) creation of an extensive ALS databank for future analysis. Patients and trained medical professionals need not be co-located, enabling more frequent monitoring of more patients from the convenience of their own homes. The goals of this study are the identification of acoustic speech features in naturalistic contexts which characterize disease progression and development of machine models which can recognize the presence and severity of the disease. We evaluated subjects from the Prize4Life Israel dataset, using a variety of frequency, spectral, and voice quality features. The dataset was generated using the ALS Mobile Analyzer, a cellphone app that collects data regarding disease progress using a self-reported ALSFRS-R questionnaire and several active tasks that measure speech and motor skills. Classification via leavefive-subjects-out cross-validation resulted in an accuracy rate of 79% (61% chance) for males and 83% (52% chance) for females.

Bulbar ALS Detection Based on Analysis of Voice Perturbation and Vibrato

“…The need for development voice and speech diagnostic tools has motivated recent work on elaboration of special acoustic analysis methods. One important aspect of current research related to neurological disease detection (such as Parkinson's and ALS) is that they targeted to use smartphones and tablets to record the voice using standard microphone in the different home environments [6,7]. Advancement in this direction can lead to creation of easy to use self-and tele-monitoring and disease detection means.…”

“…In several studies detection of ALS is performed using kinematic sensors [4,12] to model articulation and measure prosodic elements such as vowel duration or speaking rate [2]. Running speech test was used in [7,10,11] as a basis for ALS detection. In [10] representational learning approach based on convolutional neural networks (CNN) was applied for ALS detection.…”

Voice spectrum of patients diagnosed with bulbar ALS : Slope analysis in the lower energy bands

“…Em [5] surge um primeiro trabalho que sugere o uso do processamento da fala para deteção preliminar de ELA bulbar. Mais recentemente foram publicados vários trabalhos onde a análise de fala é usada para deteção de ELA com recurso a características acústicas como o jitter e o shimmer [6]- [8], sendo nestes casos usados classificadores de alto nível como discriminadores lineares e redes neuronais. As taxas de acertos rondam os 90%.…”

Voice Pathology Detection based on Analysis of Modulation Spectrum in Critical Bands

“…Приведенные графики выявляют особенности анализа, присущие рассматриваемым схемам. Схема анализа на основе ДПФ формирует сглаженное частотно-временное распределение энергии сигнала за счет относительно широкого ( [31], таких как болезнь Паркинсона [32,33], боковой амиотрофический склероз [34][35][36][37] и другие. [38].…”