A Clinical Decision Support System to Identify and Stage Breast Cancer Tumor

Abstract

Aims: An improved Clinical Decision Support System is developed to classify the tumor and identify the different stages of the breast cancer. Methodology: In this paper, we analyze and compare the performance of a developed system that takes the breast density information into account. The advantages of consideration of this breast density information will be highlighted. Our proposal is based on multi-resolution Gray Level and Local Difference (GLLD) for feature extraction. Once the descriptors are extracted, Artificial Neural Network (ANN) are used for classifying the detected masses according to their corresponding stages. Results: The accuracy of the proposed system has been verified and found that the Area Under the Curve (AUC) of 99.5% can be achieved for tumor staging when considering the information of beast density and applying the multi-resolution GLLD as texture descriptor. The proposed system may provide valuable information concerning cancer staging.