Published Paper Details:

Breast cancer develops when abnormal cells accumulate in the breasts. It was the most common cancer in women worldwide in 2020, with an estimated 2.3 million new cases being diagnosed. The aetiology of breast cancer is not yet fully understood; however, it is established that advancing age, familial history of breast cancer, genetic abnormalities, exposure to radiation, and hormonal factors are all recognised as potential risk factors for the development of this disease. The symptoms of breast cancer encompass the presence of a lump or mass in the breast, alterations in breast size or shape, skin dimpling or puckering, nipple discharge or inversion, and breast pain or tenderness. However, not all breast cancers manifest themselves in obvious ways; others are detectable only by mammography or other imaging procedures. Better patient outcomes and lower mortality rates can be achieved through early detection and precise diagnosis. Predicting breast cancer risk, recurrence, and survivability is an area where machine learning algorithms have made significant strides in recent years. This study focuses on utilising machine learning to create precise predictions regarding a range of outcomes related to breast cancer. To begin, a model is constructed to anticipate the probability of acquiring breast cancer before the onset of the disease. This is accomplished through the use of algorithms like Logistic Regression (LR), Decision Trees (DT), and Neural Networks (NN) to examine parameters including age, family history, hormone considerations, and lifestyle factors. After the model has been trained and tested on a sizable dataset of breast cancer patients and healthy individuals, a variety of metrics are used to evaluate the model's performance, including accuracy, sensitivity, specificity, and area under the receiver operating characteristic (AUC-ROC). Second, a model is constructed to foretell the likelihood of a return of breast cancer following initial clearance of the disease. This is accomplished through the use of algorithms like RF, gradient boosting, and deep learning to examine characteristics such tumour size, grade, receptor status, and treatment history. The model's performance is assessed in terms of a number of different outcomes, and it is trained and tested using data from breast cancer patients who have already had treatment and have been followed up on. Finally, a model is created to foretell how breast cancer patients will respond to treatment and whether they will survive. Support vector machines (SVM), Naive Bayes (NB), and K-nearest Neighbours(k-NN) are some of the algorithms used to analyse variables such patient demographics, tumour characteristics, and treatment history to reach this goal. Overall survival, disease-free survival, and progression-free survival are a few of the measures used to assess the model's performance once it has been trained and tested on a dataset of breast cancer patients with known outcomes. The overall goal of developing machine learning models for breast cancer prediction and survivorship is to enable earlier detection, personalised treatment planning, and improved patient outcomes, all of which have the potential to revolutionise breast cancer care.