Is there any way to make a clear or close border of region of interest in medical ultrasound images by using Matlab ?

In particular, I am working with 2D medical ultrasound images . The problem with some kinds of images is, there is no clear or close border. that will make a problem when I wanna segment the ROI by using for example Active contours method .is there any solution that will make a clear and close border before doing segmentation. Many Thanks.

Answer 1

[Image Analyst]

Use imclearborder(). Post your image is you want any help.

[Dhurgham Al-karawi]

hey .this is one example on the data which I have it .there is no close border .Many Thanks .

[Walter Roberson]

Pad the image with 0 all around to provide an artificial border.

Answer 2

[Image Analyst]

I don't know what you want to make a border around in that image. I suggest you look at the image processing literature for ultrasound segmentation papers: http://www.visionbib.com/bibliography/contentsmedical.html#Medical%20Applications,%20CAT,%20MRI,%20Ultrasound,%20Heart%20Models,%20Brain%20Models

If you want to use activecontour(), see my attached demo.

Answer 3

[Dhurgham Al-karawi]

thanks for your reply

.the purpose of close border is to work with segmentation stage .if u see the attached image ,I closed the border by the white colour just to give you an idea what I want exactly .because the most important part in may work is detecting the ROI . I have some images with clear and close border ,I did not get any problem with the segmentation stage but others as I mentioned there is no border so that will effect on the segmentation stage. many thanks .

[Image Analyst]

I didn't know if you mean close, like "nearby" or if you mean close like "to seal off". To close/seal the hole/bay you can use bwconvhull() on the binary image. If you want to draw a line in the gray scale image with the perimeter, you can use bwconvhull, and then bwperim, and burn the perimeter into the image

binaryImage = bwconvhull(binaryImage, 'union');
perimeterImage = bwperim(binaryImage) > 0;
grayImage(perimeterImage) = 255;