Detection and classification of breast cancer in mammogram images using entropy-based Fuzzy C-Means Clustering and RMCNN

Radiologists employ mammograms for the detection of breast cancer in patients, particularly as breast cancer exhibits higher incidence rates in women. Early identification of breast cancer significantly reduces the risk of mortality. Mammograms serve as a valuable imaging technique for the early detection of breast cancer. However, accurately characterizing breast cancer images poses a considerable challenge. A recent research study introduced an innovative algorithm for Mammogram Pectoral Muscle Removal, leveraging Entropy-Based Fuzzy C-means clustering and Classification using RMCNN (Root Mean Squares Convolutional Neural Network). The process involves extracting and pre-processing the input breast image from the dataset through Gaussian filtering. Subsequently, pectoral muscle removal is achieved via entropy-based fuzzy C-Means clustering, followed by DCT (Discrete Cosine Transform) and DWT (Discrete wavelet transforms) feature extraction. The Root Mean Square Value-based Convolutional Neural Network classifier effectively clusters mammography images into normal, malignant, and benign classes, achieving an impressive 99.45% accuracy far better than existing methods with accuracies of 93%, 89%, and 60%.