Classification of Breast Tumors Through Image Mining Techniques

• Lizianne P. Marques Souto
• Thiago K. L. dos Santos
• Marcelino Pereira S. Silva

DOI: https://doi.org/10.5753/sbcas.2018.3667

Resumo

In this paper a methodology and a computer-aided diagnosis system for detection of breast cancer are proposed. The approach involves Image Processing resources to extract morphological features from tumors in mammograms and Image Mining to classify them as benign or malignant. Images from BCDR repository were used for the experiments. The results showed the efficacy of the proposed method and system, which reduced the false positive and false negative rates, and allowed a more efficient decision-making process.