Time-frequency modeling and classification of pathological voices

Acoustic measures of vocal function are routinely used for the assessment of disordered voices, and for monitoring patients' progress over the course of therapy. In the paper, speech signals were decomposed using an adaptive time-frequency transform algorithm, and the signal decomposition parameters such as the octave (scale) maximum, octave mean, and frequency ratio were analyzed using a statistical pattern analysis method. A classification accuracy of 93.4% was obtained with a database of 212 speech signals (51 normal and 161 pathological cases).