Article
Automatic categorization of laryngeal images using the multiple feature sets
Search Medline for
Authors
Published: | April 22, 2008 |
---|
Outline
Text
The aim: An approach to integrating the global and local kernel-based automated analysis of vocal fold images aiming to categorize laryngeal diseases is presented in this study.
Methods: This study uses a set of 785 digital microlaryngoscopic images. The images were acquired during routine direct microlaryngoscopy employing the Moller-Wedel Universa 300 surgical microscope. The 3-CCD Elmo color video camera of 768x576 pixels was used to record the images. A rather common, clinically discriminative group of laryngeal diseases was chosen for the analysis, i.e. mass lesions of vocal folds.
This study is concerned with an automated analysis of microlaryngoscopic images aiming to categorize the images into three decision classes, namely healthy, nodular (localized thickenings): nodules, polyps, and cysts, and diffuse: papillomata, hyperplastic laryngitis with keratosis, and carcinoma mass lesions. The problem is treated as an image analysis and classification task. Aiming to obtain a comprehensive description of laryngeal images, multiple feature sets exploiting information on image color, texture, geometry, image intensity gradient direction, and frequency content are extracted. A separate support vector machine (SVM) is used to categorize features of each type into the decision classes.
Results: The final image categorization is then obtained based on the decisions provided by a committee of support vector machines. Bearing in mind a high similarity of the decision classes, the correct classification rate of over 94% obtained when testing the system on 785 microlaryngoscopic images is rather promising.