Neuroendocrinology Letters, Home
Neuroendocrinology Letters Vol. 19 Contents
Contact us
Subscription info
Neuroendocrinology Letters incl. Psychoneuroimmunology & Chronobiology


HIGHLIGHT TEXT AND CLICK TO SEARCH THE WEB

NeuroendocrinologyİLetters incl. Psychoneuroimmunology & Chronobiology,
ISSNİ0172ñ780X Copyrightİ©İ1998, 1999, 2000, 2001 NeuroendocrinologyİLetters

NEL VOL. 19 No. 3
ORIGINAL ARTICLE

1998; 19:147-157

 

Application of Neural Networks and Pattern Recognition Methods to the Evaluation of Speech Deformation Degree for Patients Surgically Treated for Larynx Cancer
by Elzbieta Reron, Ryszard Tadeusiewicz, Maciej Modrzejewski & Wieslaw Wszolek


Key words:

cancer of the larynx; voice quality; neural networks

Abstract:

The work presented concerns the automatic evaluation of pathological speech, and finding the methods of objective acoustic evaluation of the degree of speech deformation. The most interesting techniques are neural networks, which adaptation and learning ability is well known, and realization of which is now possible in terms of equipment. The objective of this presentation is to define the degree to which the technique of the neural networks can be applied in evaluation of speech deformation in a given type of pathology. The presented methodology can be directly applied in the monitoring examinations for patients after larynx oncological operations and in the future its wider use seems possible in the evaluation and optimalisation of the results of surgical treatments in many types of affliction in the larynx, pharynx, tongue, mouth and paranasal sinusis area.

__________________________________________________________
Copyright © Neuroendocrinology Letters 1998, 1999, 2000, 2001
All rights reserved. No part may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or ortherwise, without prior written permission from the Editor-in-Chief editor@nel.edu