Please use this identifier to cite or link to this item:

TitleSegmentation of bladder tumors in cystoscopy images using a MAP approach in different color spaces
Author(s)Freitas, Nuno R.
Vieira, Pedro Miguel
Lima, Estêvão Augusto Rodrigues de
Lima, C. S.
Issue date29-Mar-2017
Abstract(s)Nowadays the diagnosis of bladder lesions relies upon cystoscopic examination and depend on the interpreter's experience. State of the art of bladder tumor identification are based on 3D reconstruction, using CT images (Virtual Cystoscopy) or images where the structures are exalted with the use of pigmentations, but none uses white light cystoscopy images. Traditional cystoscopic images processing has a huge potential to improve early tumor detection and allow a more effective treatment. In this paper is described an initial approach to do segmentation of bladder cystoscopic images. This approach will be used in the future to automatically detect these types of lesions. It can be assumed that each region has a normal distribution with specific parameters, leading to the assumption that the distribution of intensities is a Gaussian Mixture Model (GMM). The most common bladder tumor type, with a cauliflower shape, appears with higher intensity than normal regions. The segmentation of these images is based on a Maximum A Posteriori (MAP) approach depending on pixel intensities of each three RGB and HSV channels, using the Expectation-Maximization (EM) algorithm to estimate the best GMM parameters. Experimental results show that the proposed method does bladder tumor segmentation in a more efficient way in RGB color space than in HSV, even in cases where the tumor shape is not well defined. Results also show that the channels with best results are the R component from RGB and the V component from HSV.
TypeConference paper
AccessRestricted access (UMinho)
Appears in Collections:ICVS - Artigos em livros de atas / Papers in proceedings
DEI - Artigos em atas de congressos internacionais

Files in This Item:
File Description SizeFormat 
  Restricted access
346,98 kBAdobe PDFView/Open

Partilhe no FacebookPartilhe no TwitterPartilhe no DeliciousPartilhe no LinkedInPartilhe no DiggAdicionar ao Google BookmarksPartilhe no MySpacePartilhe no Orkut
Exporte no formato BibTex mendeley Exporte no formato Endnote Adicione ao seu ORCID