Limits...
Optic disc segmentation by balloon snake with texture from color fundus image.

Sun J, Luan F, Wu H - Int J Biomed Imaging (2015)

Bottom Line: It uses template matching to find the region of interest (ROI).The experimental results show the superior performance of the proposed method when compared to some traditional segmentation approaches.An average segmentation dice coefficient of 94% has been obtained.

View Article: PubMed Central - PubMed

Affiliation: School of Information & Control Engineering, Shenyang Jianzhu University, Shenyang 110168, China.

ABSTRACT
A well-established method for diagnosis of glaucoma is the examination of the optic nerve head based on fundus image as glaucomatous patients tend to have larger cup-to-disc ratios. The difficulty of optic segmentation is due to the fuzzy boundaries and peripapillary atrophy (PPA). In this paper a novel method for optic nerve head segmentation is proposed. It uses template matching to find the region of interest (ROI). The method of vessel erasing in the ROI is based on PDE inpainting which will make the boundary smoother. A novel optic disc segmentation approach using image texture is explored in this paper. A cluster method based on image texture is employed before the optic disc segmentation step to remove the edge noise such as cup boundary and vessels. We replace image force in the snake with image texture and the initial contour of the balloon snake is inside the optic disc to avoid the PPA. The experimental results show the superior performance of the proposed method when compared to some traditional segmentation approaches. An average segmentation dice coefficient of 94% has been obtained.

No MeSH data available.


Related in: MedlinePlus

Optic disc template [12].
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4378594&req=5

fig2: Optic disc template [12].

Mentions: We use the template matching [9] to find the estimated disc region and then apply a circle Hough transform to correct the center and find an estimated disc radius. Since the appearance of the OD may vary significantly due to retinal pathologies, but the optic disc is the brightest feature of the normal fundus and shape of the optic disc is approximately vertically slightly oval (elliptical), a lot of templates have been proposed in [12] to find the optic disc. In this paper, we adopt the template from Lowell et al. [9] as shown in Figure 2. In our dataset the optic disc width is ranging from 260 to 380 pixels' length. So a template with size of 401∗401 is used in our experiments. Then the Pearson correlation coefficient is used to measure the correlation between the intensity channel from HSI color space subimage and the template in general:(1)cij=∑x,yfx,y−fmtx−i,y−j−tm∑x,yfx,y−fm2∑x,ytx−i,y−j−tm2,where tm is the mean intensity values of the template, which need to be calculated only once. fm is the mean intensity values of the image region covered by the template. The location with the maximum value is selected as the optic disc location. The method is robust to small exudate and can fast locate the optic disc. In the next step a circle Hough transform is used to update the center and find a disc radius.


Optic disc segmentation by balloon snake with texture from color fundus image.

Sun J, Luan F, Wu H - Int J Biomed Imaging (2015)

Optic disc template [12].
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4378594&req=5

fig2: Optic disc template [12].
Mentions: We use the template matching [9] to find the estimated disc region and then apply a circle Hough transform to correct the center and find an estimated disc radius. Since the appearance of the OD may vary significantly due to retinal pathologies, but the optic disc is the brightest feature of the normal fundus and shape of the optic disc is approximately vertically slightly oval (elliptical), a lot of templates have been proposed in [12] to find the optic disc. In this paper, we adopt the template from Lowell et al. [9] as shown in Figure 2. In our dataset the optic disc width is ranging from 260 to 380 pixels' length. So a template with size of 401∗401 is used in our experiments. Then the Pearson correlation coefficient is used to measure the correlation between the intensity channel from HSI color space subimage and the template in general:(1)cij=∑x,yfx,y−fmtx−i,y−j−tm∑x,yfx,y−fm2∑x,ytx−i,y−j−tm2,where tm is the mean intensity values of the template, which need to be calculated only once. fm is the mean intensity values of the image region covered by the template. The location with the maximum value is selected as the optic disc location. The method is robust to small exudate and can fast locate the optic disc. In the next step a circle Hough transform is used to update the center and find a disc radius.

Bottom Line: It uses template matching to find the region of interest (ROI).The experimental results show the superior performance of the proposed method when compared to some traditional segmentation approaches.An average segmentation dice coefficient of 94% has been obtained.

View Article: PubMed Central - PubMed

Affiliation: School of Information & Control Engineering, Shenyang Jianzhu University, Shenyang 110168, China.

ABSTRACT
A well-established method for diagnosis of glaucoma is the examination of the optic nerve head based on fundus image as glaucomatous patients tend to have larger cup-to-disc ratios. The difficulty of optic segmentation is due to the fuzzy boundaries and peripapillary atrophy (PPA). In this paper a novel method for optic nerve head segmentation is proposed. It uses template matching to find the region of interest (ROI). The method of vessel erasing in the ROI is based on PDE inpainting which will make the boundary smoother. A novel optic disc segmentation approach using image texture is explored in this paper. A cluster method based on image texture is employed before the optic disc segmentation step to remove the edge noise such as cup boundary and vessels. We replace image force in the snake with image texture and the initial contour of the balloon snake is inside the optic disc to avoid the PPA. The experimental results show the superior performance of the proposed method when compared to some traditional segmentation approaches. An average segmentation dice coefficient of 94% has been obtained.

No MeSH data available.


Related in: MedlinePlus