Characterizing vocal tremor in progressive neurological diseases via automated acoustic analyses

Abstract Objective Voice tremor represents a common but frequently overlooked clinical feature of neurological disease. Therefore, we aimed to quantitatively and objectively assess the characteristics of voice tremor in a large sample of patients with various progressive neurological diseases. Methods Voice samples were acquired from 240 patients with neurological disease and 40 healthy controls. The robust automated method was designed, allowing precise tracking of multiple tremor frequencies and distinguish pathological from the physiological tremor. Results Abnormal tremor was revealed in Huntington’s disease (65%), essential tremor (50%), multiple system atrophy (40%), cerebellar ataxia (40%), amyotrophic lateral sclerosis (40%), progressive supranuclear palsy (25%), Parkinson’s disease (20%), cervical dystonia (10%), and multiple sclerosis (8%) but not in controls. Low-frequency voice tremor ( Conclusions Careful estimation of vocal tremor may help with accurate diagnosis and tailored treatment. Significance This study provides (i) more insights into the pathophysiology of vocal tremor in a wide range of neurological diseases and (ii) an accurate method for estimation of vocal tremor suitable for clinical practice.