Automatic detection of laryngeal pathologies in running speech based on the HMM transformation of the nonlinear dynamics

This work describes a novel system for characterizing Laryngeal Pathologies using nonlinear dynamics, considering different complexity measures that are mainly based on the analysis of the time delay embedded space. The model is done by a kernel applied on Hidden Markov Model and decision of the Laryngeal pathology/control detection is performed by Support Vector Machine. Our system reaches accuracy up to 98.21%, improving the current reported results in the state of the art in the automatic classification of pathological speech signals (running speech) and showing the robustness of this proposal