Identification Of Patients At The Risk Of Lewy Body Diseases Based On Acoustic Analysis Of Speech

Abstract

Lewy body diseases (LBDs) is a group of neurodegenerative diseases that consists of Parkinson’s disease and dementia with Lewy bodies, that are generally very crucial to be diagnosed in their prodromal state. In the frame of this study we proposed a multivariate logistic regression model that identifies people in a high risk of LBDs based on their articulatory and prosodic characteristics. More specifically, the model has 80 % specificity and 85 % sensitivity based on quantification of rigidity of tongue/jaw, monoloudness, and inappropriate pausing.