This repository presents an interface designed to assist medical professionals in the analysis of mammograms, aimed at early breast cancer detection. The interface, built using QT5 and Python, enables users to load mammographic images, select regions of interest, and calculate relevant texture metrics (homogeneity, entropy, dissimilarity, correlation, energy, and contrast). It also provides a visual segmentation of dense tissue regions and predicts breast tissue density using Random Forest and BIRADs classification. The tool's accuracy is 86% in classifying breast tissue density.
Development team: Cerretini, Costa, Espinola and Galansky.