Voice Pathology Detector: desenvolvimento de uma aplicação móvel para detecção de patologias de voz.

Abstract

In the last decade, several research studies have been developed with the aim of classifying and identifying voice pathologies. In this context, the existence of a widely accessible tool to provide non-invasive pre-diagnosis for voice disease detection would be highly relevant in motivating users to seek medical assistance. This work proposes the development of a mobile application that classifies healthy and unhealthy voices in a binary manner. Supervised machine learning techniques are employed, along with a voice database containing healthy voice signals and signals affected by some pathologies, for training and validation purposes. The resulting application has an intuitive use and presents a relevant social impact by helping reduce the gap between doctors and individuals with voice pathologies. By leveraging this application, users will have an accessible means to assess their voice health, encouraging timely medical intervention.