Limits...
Computer-assisted segmentation of videocapsule images using alpha-divergence-based active contour in the framework of intestinal pathologies detection.

Meziou L, Histace A, Precioso F, Romain O, Dray X, Granado B, Matuszewski BJ - Int J Biomed Imaging (2014)

Bottom Line: Nevertheless, the systematic postexamination by the specialist of the 50,000 (for the small bowel) to 150,000 images (for the colon) of a complete acquisition using WCE remains time-consuming and challenging due to the poor quality of WCE images.In this paper, a semiautomatic segmentation for analysis of WCE images is proposed.Results of segmentation using the proposed approach are shown on different types of real-case examinations, from (multi)polyp(s) segmentation, to radiation enteritis delineation.

View Article: PubMed Central - PubMed

Affiliation: ETIS, Université de Cergy-Pontoise, ENSEA, CNRS, 95014 Cergy-Pontoise Cedex, France.

ABSTRACT
Visualization of the entire length of the gastrointestinal tract through natural orifices is a challenge for endoscopists. Videoendoscopy is currently the "gold standard" technique for diagnosis of different pathologies of the intestinal tract. Wireless capsule endoscopy (WCE) has been developed in the 1990s as an alternative to videoendoscopy to allow direct examination of the gastrointestinal tract without any need for sedation. Nevertheless, the systematic postexamination by the specialist of the 50,000 (for the small bowel) to 150,000 images (for the colon) of a complete acquisition using WCE remains time-consuming and challenging due to the poor quality of WCE images. In this paper, a semiautomatic segmentation for analysis of WCE images is proposed. Based on active contour segmentation, the proposed method introduces alpha-divergences, a flexible statistical similarity measure that gives a real flexibility to different types of gastrointestinal pathologies. Results of segmentation using the proposed approach are shown on different types of real-case examinations, from (multi)polyp(s) segmentation, to radiation enteritis delineation.

No MeSH data available.


Related in: MedlinePlus

Three examples of the proposed boosting-based polyp detection approach in [2].
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4281406&req=5

fig3: Three examples of the proposed boosting-based polyp detection approach in [2].

Mentions: Regarding the first issue, in [2], we proposed an automatic boosting-based detection approach for abnormal region of interest (ROI) identification, with a particular focus on polyp detection. A database of 500 polyps and 1200 nonpolyps was used to learn by boosting the optimal classifier in terms of false positive (FP) and true positive (TP) rates. In that study, texture parameters were used (computed from the cooccurrence matrix) and a global detection rate of 94% was reached with a FP rate of 4%. That work, achieved on classic videoendoscopic images, is currently being extended to other types of structure than polyps and to WCE images, for which the image database is far more difficult to build because of limited access to raw images data. Figure 3 displays three examples of obtained results. In each case, nonbold squares show candidate regions before the classifying step, whereas the bold square shows the region eventually identified as a polyp using the boosted classifier. It is important to highlight here that a resolution of only 5 bits (32 grey-levels) was used to compute the cooccurrence matrices from which the Haralick's parameters are then computed. It is shown that, even for low-resolution WCE images (with respect to classic resolution of videocolonoscopy images), the method remains successful.


Computer-assisted segmentation of videocapsule images using alpha-divergence-based active contour in the framework of intestinal pathologies detection.

Meziou L, Histace A, Precioso F, Romain O, Dray X, Granado B, Matuszewski BJ - Int J Biomed Imaging (2014)

Three examples of the proposed boosting-based polyp detection approach in [2].
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4281406&req=5

fig3: Three examples of the proposed boosting-based polyp detection approach in [2].
Mentions: Regarding the first issue, in [2], we proposed an automatic boosting-based detection approach for abnormal region of interest (ROI) identification, with a particular focus on polyp detection. A database of 500 polyps and 1200 nonpolyps was used to learn by boosting the optimal classifier in terms of false positive (FP) and true positive (TP) rates. In that study, texture parameters were used (computed from the cooccurrence matrix) and a global detection rate of 94% was reached with a FP rate of 4%. That work, achieved on classic videoendoscopic images, is currently being extended to other types of structure than polyps and to WCE images, for which the image database is far more difficult to build because of limited access to raw images data. Figure 3 displays three examples of obtained results. In each case, nonbold squares show candidate regions before the classifying step, whereas the bold square shows the region eventually identified as a polyp using the boosted classifier. It is important to highlight here that a resolution of only 5 bits (32 grey-levels) was used to compute the cooccurrence matrices from which the Haralick's parameters are then computed. It is shown that, even for low-resolution WCE images (with respect to classic resolution of videocolonoscopy images), the method remains successful.

Bottom Line: Nevertheless, the systematic postexamination by the specialist of the 50,000 (for the small bowel) to 150,000 images (for the colon) of a complete acquisition using WCE remains time-consuming and challenging due to the poor quality of WCE images.In this paper, a semiautomatic segmentation for analysis of WCE images is proposed.Results of segmentation using the proposed approach are shown on different types of real-case examinations, from (multi)polyp(s) segmentation, to radiation enteritis delineation.

View Article: PubMed Central - PubMed

Affiliation: ETIS, Université de Cergy-Pontoise, ENSEA, CNRS, 95014 Cergy-Pontoise Cedex, France.

ABSTRACT
Visualization of the entire length of the gastrointestinal tract through natural orifices is a challenge for endoscopists. Videoendoscopy is currently the "gold standard" technique for diagnosis of different pathologies of the intestinal tract. Wireless capsule endoscopy (WCE) has been developed in the 1990s as an alternative to videoendoscopy to allow direct examination of the gastrointestinal tract without any need for sedation. Nevertheless, the systematic postexamination by the specialist of the 50,000 (for the small bowel) to 150,000 images (for the colon) of a complete acquisition using WCE remains time-consuming and challenging due to the poor quality of WCE images. In this paper, a semiautomatic segmentation for analysis of WCE images is proposed. Based on active contour segmentation, the proposed method introduces alpha-divergences, a flexible statistical similarity measure that gives a real flexibility to different types of gastrointestinal pathologies. Results of segmentation using the proposed approach are shown on different types of real-case examinations, from (multi)polyp(s) segmentation, to radiation enteritis delineation.

No MeSH data available.


Related in: MedlinePlus