Region Merging for Lumen Detection

Figure 9.23 shows how the regions with multiple classes are merged. We will discuss the region merging strategy a little differently for the real data analysis,

Figure 9.22: Detection and identification of lumen. Input image is a classified image with multiple classes inside the lumens. Given the number of classes K and the region of interest (ROI) of each region, the appropriate classes are merged and the image is binarized. The detected lumens are then identified using connected component analysis (CCA), and the left lumen and right lumen are identified.

Figure 9.22: Detection and identification of lumen. Input image is a classified image with multiple classes inside the lumens. Given the number of classes K and the region of interest (ROI) of each region, the appropriate classes are merged and the image is binarized. The detected lumens are then identified using connected component analysis (CCA), and the left lumen and right lumen are identified.

due to the bifurcations in the arteries of the plaqued vessels (see sections 9.6.1 and 9.6.2). Figure 9.23 illustrates the region merging algorithm. The input image has lumens which have one, two, or more classes. If the number of classes in the ROI is one class, then that class is selected; if two classes are in the ROI, then the minimum class is selected; and if there are three or more classes in the ROI, then the minimum two classes are selected. The selected classes are merged by assigning all the pixels of the selected classes one level value. This process results in the binarization of the left and right lumens.

The binary region labeling process is shown in Fig. 9.24. The process uses the CCA approach of top to bottom and left to right. Input is an image in which the lumen regions are binarized. The CCA first labels the image from the top to the bottom, and then from the left to the right. The result is an image that is labeled from the left to the right.

ID assignment process of the CCA for each pixel is shown in Fig. 9.25. In the CCA, in the input binary image, each white pixel is assigned a unique ID. The label propagation process then results in connected components. The propagation of

Figure 9.23: Region detection: region merging algorithm. The input image has lumens that have 1, 2, or more classes. If the number of classes in the ROI is one class, then that class is selected; if two classes are in the ROI, then the minimum class is selected; and if there are three or more classes in the ROI, then the minimum two classes are selected. The selected classes are merged by assigning all the pixels of the selected classes one level value. This process results in the left and right lumen being binarized.

Figure 9.23: Region detection: region merging algorithm. The input image has lumens that have 1, 2, or more classes. If the number of classes in the ROI is one class, then that class is selected; if two classes are in the ROI, then the minimum class is selected; and if there are three or more classes in the ROI, then the minimum two classes are selected. The selected classes are merged by assigning all the pixels of the selected classes one level value. This process results in the left and right lumen being binarized.

Figure 9.24: Region identification using connected component analysis (CCA). Input is an image in which the lumen binary regions are detected. The CCA first labels the image from the top to the bottom, and then from the left to the right. The result is an image that is labeled from the left to the right.

Figure 9.24: Region identification using connected component analysis (CCA). Input is an image in which the lumen binary regions are detected. The CCA first labels the image from the top to the bottom, and then from the left to the right. The result is an image that is labeled from the left to the right.

Binary Image

i

i

i

i

i

i

i

i

Assign each white pixel an ID

i

i

i

i

i

Each white pixel has unique ID

Binary Image

Connected Components

Figure 9.25: Region identification: ID assignment. In the connected component analysis (CCA), in the input binary image each white pixel is assigned a unique ID. Then the label propagation process results in connected components.

Connected Components

i

2

s

4

s

e

7

a

iG

ii

is

i4

is

Assigning unique IDs

Figure 9.25: Region identification: ID assignment. In the connected component analysis (CCA), in the input binary image each white pixel is assigned a unique ID. Then the label propagation process results in connected components.

the region from left to right is shown in Fig. 9.26. This is the first pass of the label-propagation process. Every row of the image is scanned from top to bottom, left to right, pixel by pixel. If the pixel has an ID, then pixels to the left and above of the pixel are checked for IDs, and if either one has an ID, then the pixel's value is reassigned to that of the lowest among the neigbor pixels and the pixel. This processed is repeated for all pixels in the row and in all rows. The result is a binary image with some label propagation. The propagation of the region from top to bottom is shown in Fig. 9.27. This is the second pass of the label-propagation process. Every row of the image is scanned from bottom to top, right to left, pixel by pixel. If the pixel has an ID, then pixels to the right and below of the pixel are checked for IDs, and if either one has an ID, then the pixel's value is reassigned to that of the lowest among the neigbor pixels and the pixel. This processed is repeated for all pixels in the row and in all rows. The result is a binary image with some label propagation. Finally, the region assignment is summarized in Fig. 9.28. The top left image is a binary image with a value of 1 assigned to each of the white pixels. Each of the white pixels are assigned a unique value in the top-right image. The left to right and top to bottom label propagation propagates the labels of value 1 and 3, and the result is the

Figure 9.26: Region identification: propagation. This is the first pass of the label propagation process. Given the bianary image having unique IDs for each white pixel, every row of the image is scanned from top to bottom, left to right, pixel by pixel. If the pixel has an ID, then pixels to the left and above of the pixel are checked for IDs, and if either one has an ID, then the pixel's value is reassigned to that of the lowest among the neigbor pixels and the pixel. This processed is repeated for all pixels in the row all rows. The result is a binary image with reassigned pixel values.

Figure 9.26: Region identification: propagation. This is the first pass of the label propagation process. Given the bianary image having unique IDs for each white pixel, every row of the image is scanned from top to bottom, left to right, pixel by pixel. If the pixel has an ID, then pixels to the left and above of the pixel are checked for IDs, and if either one has an ID, then the pixel's value is reassigned to that of the lowest among the neigbor pixels and the pixel. This processed is repeated for all pixels in the row all rows. The result is a binary image with reassigned pixel values.

bottom-left image. Then, the right to left and bottom to top label propagation propagates the label value of 1 to the pixels having a label value of 3. The result is the bottom-right image, in which the connected white pixels have all the same label values of 1. This is the basic algorithm of the process; the CCA we used uses look-up tables in order to efficiently assign regions in two passes. The results on CCA on a binary image with 4 lumens are shown in Fig. 9.29. The input image has the lumens detected, but they are all of the same color. CCA identifies the lumens by labeling each with a different color. The process to generate a color image is shown in Fig. 9.30. The first input is a gray scale image. The second input is the ideal boundary image. This image is dilated and converted to a red color, resulting in a red ideal boundary image. The third input is the estimated boundary image. This image is dilated and converted to a green color, resulting

Figure 9.27: Region identification: propagation. This is the second pass of the label propagation process. Given the binary image having unique IDs for each white pixel, every row of the image is scanned from bottom to top, right to left, pixel by pixel. If the pixel has an ID, then pixels to the right and below of the pixel are checked for IDs, and if either one has an ID, then the pixel's value is reassigned to that of the lowest among the neigbor pixels and the pixel. This processed is repeated for all pixels in the row and all rows. The result is a binary image with reassigned pixel values.

Figure 9.27: Region identification: propagation. This is the second pass of the label propagation process. Given the binary image having unique IDs for each white pixel, every row of the image is scanned from bottom to top, right to left, pixel by pixel. If the pixel has an ID, then pixels to the right and below of the pixel are checked for IDs, and if either one has an ID, then the pixel's value is reassigned to that of the lowest among the neigbor pixels and the pixel. This processed is repeated for all pixels in the row and all rows. The result is a binary image with reassigned pixel values.

in green estimated boundary image. These three images are fused to produce a color overlay image.

0 0

Post a comment