Développement d’un Système D’aide au Diagnostic en Imagerie Mammographique

Abstract

Mammography is the most widely used imaging technique for detecting tumors in their earliest stages. The presence of calcifications in mammographic images is of particular interest for the early detection of breast cancer. We have implemented a system based on mathematical morphologies, microcalcification classification with Resnet50 and segmentation using Otsu filtering and thresholding. The use of Resnet50 for microcalcification classification in mammographic images enables complex features to be learned from the images, which can be crucial for the accurate detection of microcalcifications. Segmentation of microcalcifications by Otsu thresholding is a method used in medical image processing, and is often used for two reasons: simplicity and adaptability.