0

I am trying to normalize MR image. There is a negative value in the MR image. So the MR image was normalized using the Gaussian method, resulting in a negative area. But i don't want to get negative area.

My question: What is the normalization method so that there is no negative value?

Thanks in advance

lv9999
  • 5
  • 2
  • To set the smallest value to zero, find the smallest value and subtract that from everything, `x[i,j] = x[i,j] - min(x)` for all pixels. Then you can, say, divide everything by the max to get it in the range 0-1. – tom10 Feb 09 '18 at 03:03

1 Answers1

1

Without more information about your source code and the packages you're using, this is really more of a data science question than a python question.

To answer your question, a more than satisfactory method in most circumstances it min-max scaling. Simply normalize each coordinate of your images between 0 and 1. Whether or not that is appropriate for your circumstance depends on what you intend to do next with that data.

Hans Musgrave
  • 6,613
  • 1
  • 18
  • 37