ABSTRACT
Breast cancer (BC) is a common cancers worldwide in addition early detection is serious for improving patient outcomes. Conversely, detecting BC in mammography images ruins a challenging task. Therefore, in this paper an innovative methodology for automated BC detection is proposed using an enhanced VGG-16 model. At first, Mean Filter (MF) is employed for pre-processing, reducing noise as well as enhancing image quality for more precise analysis. Segmentation is then done using a fractal threshold method to isolate regions of interest that are important for detection. Feature extraction is supported out using Local Binary Patterns (LBP), which capture the texture as well as patterns of the segmented regions. These features are then classified using a modified VGG-16 Convolutional Neural Network (CNN), aimed to improve BC detection accuracy. The proposed approach employed using python, achieves an accuracy of 92% and precision of 92%, aiding medical professionals in making more accurate and early diagnoses, thereby enhancing patient prognosis and decision making in BC detection.
