Automated neural network approaches combined with chaotic grey wolf and efficient black widow optimization algorithms for breast cancer detection from mammogram images

Abstract

newline India has the dubious distinction of being in the second position worldwide in cancer deaths. Breast cancer (Carcinoma) is a very hazardous disease that affects mainly women and leads to high mortality rate. Early detection of this disease helps to reduce mastectomy and even fatality as a result of the disease. newlineMammography, the best digital screening abnormality detection tool, helps doctors to diagnose breast cancer and also aids in mammogram screening for abnormalities by radiologists to detect breast carcinoma at an early stage. For this, researchers have been continuously exploring effective algorithms to nullify complexities and increase accuracy. Despite the fact that the mammogram analysis is found to be quite successful, the accuracy rate is not satisfactory due to false positive results. Hence, in order to ensure a foolproof diagnosis, this research has postulated and developed two novel techniques utilizing Recurrent Neural Network and Pulse Coupled Neural Network by combining them with unique optimisation techniques.