Artificial Neural Network Based Effective Detection of Breast Cancer By Using Mammogram Data

Progressive MS lesion formation is a primary cause of cognitive decline and physical impairment, making MS one of the major disorders. The rate of disease progression and the efficacy of treatments depend critically on having a quick and accurate method for assessing the number and size of MS lesions in the brain. The position, size, morphological variation, intensity resemblance with normal brain tissues, and inter-subject anatomical variance of MS lesions in brain MRI make precise diagnosis, characterisation, and quantification exceedingly challenging. In this paper, we explored a technique for detecting and segmenting MS lesions in which the fundamental processes are background creation and binarization with a global threshold. To link the normal tissues close to the margins, we perform a three-phase level set and then add a third-phase segmented region using a contour of the brain. To create an adaptable setting, we must first eliminate lesions from the brain, saving only the brain's largest linked area and the corpus callosum. Selecting the threshold using entropy and standard deviation via the binarization approach is followed by non-gamut image enhancement. The proposed method detects lesions in the same locations as a radiologist and with the same precision in terms of size and number. As a result of its flexibility, the suggested method has a range of applications in clinical practise for identifying MS lesions in MRI. When compared to more contemporary methods, the proposed method has a smaller margin of error and higher accuracy.