DIGITAL IMAGE PROCESSING APPLIED TO MAMMOGRAPHY: An approach based on segmentation, enhancement, and lesion marking techniques

This study explores digital image processing techniques to support lesion detection in mammograms, using the INbreast and a private clinical dataset. Preprocessing included breast region cropping, intensity normalization, and contrast enhancement with CLAHE. Bounding boxes were created over suspicious areas using morphological criteria. The results show that these methods effectively prepare images for medical interpretation or computer-aided diagnosis, enhancing accuracy and reducing subjectivity in breast cancer evaluation.