...
- Conversion from DICOM file to PNG.
- 8 bits conversion, as Keras does not support 12 or 16 bits images.
- Background removal, the images are cropped at the breast skin edge line.
- Pectoral muscle removal in MLO projections. As it is dense tissue, it could be a source of confusion for a density classifier. It was segmented and replaced with the mean gray value of the breast
Some examples of these preprocessed mammographic images are available as test images on GitHub repository:
The aim of this tutorial is to share our knowledge to allow you to use your own dataset, which could be fundamental to evaluate the robustness of the model.
...