The cost of developing a cost function

Post date: Jul 15, 2009 12:46:14 AM

When i began this process of image registration of vertebral bodies, it seemed that the most difficult part would be to come up with a new cost function for comparing the CT generated image with the fluoro image. It turned out that this was not nearly as difficult as coming up with a method for actually using this new cost function and assessing its success (or failure). More on my registration tool later...