Speech Pathologies - Biophysically-informed Signal Processing
The linear source-filter theory of voice production is a useful approximation, but the simple assumptions
do not readily hold for real signals. Extensive modelling and experimentation have shown that speech
signals are not linear [1], and that vocal pathologies often
lead to apparently chaotic behaviour. Many of these common vocal pathologies arise due to accident,
disease, misuse of the voice, or surgery affecting the vocal folds, having a profound impact on social
and professional lives of patients.
Existing methods for automatic detection and measurement of vocal pathologies generally use classical,
linear signal processing techniques such as Fourier analysis. Due to the mismatch between the true
dynamics and the linear model, existing techniques lack internal consistency. The difficulty is these
biomedical signals exhibit very complex dynamics, which do not have sparse linear representations. We have
developed a simple nonlinear approach to online acoustic speech pathology detection
for automatic screening purposes [3-4]. It uses two nonlinear measures,
based parsimoniously upon the biophysics of speech production,
combined with a subsequent, simple classification technique.
This method achieves an overall normal/pathological detection performance of 91.8% (see figure above),
which outperforms all combinations of existing techniques. This demonstrates
that nonlinear approaches to speech pathology detection, informed by biophysics, can be both simple and
robust, and are amenable to implementation as online algorithms.
For more details, please contact Max Little.
.[1] M.A. Little, P.E. McSharry, I.M. Moroz, S.J. Roberts (2006), Testing the assumptions of linear prediction analysis in normal vowels. Journal of the Acoustical Society of America, 119(1): pp. 549-558.
[2] J. Hanquinet, G. Francis, J. Schoentgen (2005), Synthesis of disordered voices in Proceedings of 3rd International Conference on Non-Linear Speech Processing, NOLISP'05: Barcelona, Spain. pp. 168-173.
[3] M. Little, P. McSharry, I. Moroz, S. Roberts (2006), Nonlinear, Biophysically-Informed Speech Pathology Detection in 2006 IEEE International Conference on Acoustics, Speech and Signal Processing, 2006. ICASSP 2006 Proceedings.: Toulouse, France. pp. II-1080-II-1083.
[4] M.A. Little, P.E. McSharry, S.J. Roberts, D.A. Costello, I.M. Moroz (2007), Exploiting nonlinear recurrence and fractal scaling properties for voice disorder detection. Biomed Eng Online, 6: pp. 23.