Department of Image Processing

Along with the refinement of diagnostic and therapeutic techniques in phoniatrics and phonosurgery, the need for objective methods of vocal folds cycle evaluation, and its pathological or post-treatment changes, has gained more and more relevance.
This work concerns the segmentation of videokymographic images using snake algorithm. Two snakes are implemented, one for the left vocal fold and the other one for the right vocal fold. Both snakes are represented as a linked list, and a first routine is executed to provide initial data for the snakes. After the segmentation, some parameters are extracted to give clinicians an help for pathology classification and diagnosis.
To help the user in handing the snake’s parameters and patients database, we also developed a fast and simple user interface for the proposed tool. The interface allows for plotting the parameters tracking from a set of frames selected by the user. Storage and retrieval of patient’s data is also implemented.