Using Automatic Voice Analysis and Machine Learning to Screen for Swallowing Disorders in Adults with Neurological Conditions

What this Trial Is About

This research aims to develop a screening tool for oropharyngeal dysphagia, a swallowing disorder common in neurological patients such as those with Parkinsons disease, stroke, or amyotrophic lateral sclerosis. Dysphagia can lead to serious complications like malnutrition, dehydration, and pneumonia, increasing hospital stays and healthcare costs. Currently, there are no fast and effective screening methods available outside specialized clinical settings. Participants in this study perform various voice tasks during a single session, including sustained phonation, rapid syllable repetition, standardized sentences, and free speech. The recorded voice signals are analyzed using machine learning algorithms to identify changes related to swallowing difficulties. The study includes patients with neurological disorders both with and without dysphagia, as well as healthy individuals, to help develop the classification algorithm. During the study visit, participants provide voice recordings and clinical and medical history information. Researchers use these data to extract acoustic features for the machine learning model. The main goal is to create an algorithm that can quickly and non-invasively screen for swallowing disorders in neurological patients. The study is observational and involves a single recording session with no additional treatments or interventions.

Automatic Voice Analysis for Dysphagia Screening in Neurological Patients