Breast cancer detection using optimized hidden convolutional neural network

Abstract

Breast cancer is the deadlier ailment, and it is the vital cause for augmented death rates in women. Mammography is the principal method for diagnosing breast cancer. Currently, diagnosing breast cancer using mammogram images is a versatile task. Here, the proposed system develops a novel scheme for detecting breast cancer. Initially, erosion and dilation are done during pre-processing. Then the features, including Local Binary Pattern (LBP) and Gray Level Run-Length Matrix (GLRM), are extracted. Further, to resolve the issue of dimensionality, the features are chosen optimally via a new algorithm termed Lion Integrated FireFly Algorithm (LI-FF). Then, an optimized CNN is introduced in the detection phase that portrays the detected results precisely. Remarkably, the same LI-FF model fine-tuned CNN's weights and hidden neuron counts. At last, the superiority of the developed approach is proved on various measures.