BXrayBLR: Experimental Analysis & Evaluation of the Performance for LR Model on BC Prediction

In the recent two decades, cancer has been the leading cause of death around the globe. Breast Cancer (BC) is the second-highest, causing a maximum number of deaths globally these days. This, comprises pharmaceuticals, medical and surgical supplies, medical gadgets and equipment, and other products required to assist doctors, nurses, and other healthcare individuals. This work aims to address the requirement for BC specimen collection within healthcare organizations, as well as to predict BC before it happens. Firstly, we have carried out a methodology to determine the necessary steps for predicting BC. Secondly, breast screening and testing using X-ray techniques. Thirdly, a model called logistic regression (LR) to implement on breast mammograms to predict the BC disease. In the end, performance for this model has been measured by using precision, recall, f1score, and accuracy. Finally, it was found that the LR model achieves an accuracy of 97.36 % while detecting BC from breast mammograms. However, an ROC (Receiver Operator Characteristic) curve has been plotted, representing the false positive rate (FPR) and true positive rate (TPR) on the x and y-axis respectively. Thus, defining the LR model predicts the BC closer to ‘1’, which might be a greater advantage for the doctors.