Medical Decision Support Systems for Breast Cancer Diagnosis using Ensemble and Deep Learning

Abstract

Breast cancer grows in the breast tissue and then (like most cancers) spreads to newlineother areas of the body. In keeping with the numerous reports and according to global newlinestatistics, this is the second most commonly occurring malignancy overall and the most newlinepredominant, most frequently diagnosed, as well as themost invasive cancer among newlinewomen in the world. It can only be treated and controlled if it is identified at an earlier newlinelevel. Therefore, to significantly increase survival rates and the probability of recovery, newlineearly screening, accurate detection, and the right diagnosis of breast cancer are critically newlinenecessary. It also improves the chance of making the proper decision on an effective newlinemedication strategy. Computer-aided intelligent and expert diagnostic systems, built on newlinesoft computing and machine learning methods, are essential (vital) tools for evaluating newlineand predicting breast cancer and can assist oncologists (medical professionals) in the newlinedecision-making process. The medical decision support system aims to support newlinehealthcare teams and staff by examining electronic health records of patients and newlineemploys clinical knowledge on patient records to increase the efficiency of prediction. newlineThese systems help healthcare practitioners in making more precise decisions based on newlinevast amounts of complex clinical data. In the past, diverse medical decision support newlinesystems have been explored to diagnose different diseases. However, the possibilities newlinefor further improvement of diagnostic performance may be explored through new newlineand alternative scientific design methodologies or approaches. Therefore, with the newlineinsights of various capabilities and issues persisting with the traditional classification newlineapproaches, the ensemble and the deep neural network-based frameworks to classify newlinebreast cancer have been proposed and analysed in this thesis.