Voice cloud platform for medical applications

Abstract

The usability and health of a person's voice has dire impact on the person's quality of life. Pathological issues that may exist on a person's voice often cannot be detected by a regular listener. Medical attention from a professional may be necessary to detect vocal pathologies. Analysis of the patients complaints and a perceptual evaluation performed by a doctor is one of the most common ways to diagnose a vocal condition. This method can be invasive, time consuming and expensive. Features of the voice can be extracted and utilized in a computer environment to make the same diagnosis which may increase the speed and accuracy of the diagnosis and decrease the cost. In this paper, a cloud application which collects vocal data in a database is proposed. With data mining and machine learning methods, a new tool has been developed to detect and diagnose vocal anomalies in patients. The effectiveness of the suggested platform has been demonstrated with a pathological detection and recognition application running in the server.