How can I find the region of Interest in Gray scal images(Medical images)?

Actually I am working on DICOM images and my aim to find the region of interest in gray sclae images then apply lossless compression on that region of interest and rest of the part will be compress by lossy technique.So I need to find out the suitable way for region of interest.

Answer 1

[Image Analyst]

I'd use image processing / image analysis. MATLAB has an Image Processing Toolbox to help you. Then do some research on algorithms that work for your type of images, say here: http://iris.usc.edu/Vision-Notes/bibliography/contents.html. Pick one and code it up. That's about all I can say, given what you've given us to work with. Good luck!

Answer 2

[bethel o]

You can use edge detectors and apply a threshold. BW = edge(I,'sobel',THRESH). Where 'I' is your intensity image, 'sobel' is the method and 'THRESH' is you threshold. There many other techniques like 'snakes' and curve fitting but they are basically the implemented by manipulating the intensity values of intensity image. A 'Snake' known as GVF(Gradiebt Vector Force) is probably your best option because it can draw a curve arround the region of interest and you should be able to find matlab code for it easily incase you cannot be bordered writing you owm code.

[bethel o]

The matlab C-code is here, http://www.nitrc.org/frs/download.php/1890/gvfc.zip. You might need to compile a mex file from the C-code if it is not already done. Also check out http://www.iacl.ece.jhu.edu/static/gvf/ for lots of resources.

Good luck

Bethel