Search code examples
pythonpython-imaging-library

Image processing on grayscale medical image

I have been looking into PIL to perform image processing on grayscale medical images (on the breasts) so the microcalcification clusters can be seen more vividly. So far here are my current findings:

(1) Original (2) With invert, auto contrast and posterise applied (3) The yellow circled areas are the location of the clusters So I am wondering, is there a better method of image processing for this kind of images to highlight the calcification clusters? Since I will need to use them for generating graphs to show their distributional patterns later on. Any suggestions would be much appreciated.

Solution

  • I'd have a look at scikit-image. It's a great library for image processing of this kind. The first link is to the documentation, the second is to a specific page about adaptive histogram equalization that might be useful for you.

    https://scikit-image.org/

    http://scikit-image.org/docs/dev/auto_examples/color_exposure/plot_equalize.html