Data-Driven Breast Cancer Detection: A Review of Machine Learning, Deep Learning, and Multimodal AI Approaches

Breast cancer is a critical and life threatening disease among female individuals in the globe and the early diagnosis of this disease is quite important in enhancing the survival rates and treatment outcomes. The paper will provide a description of machine learning and deep learning algorithms of breast cancer diagnosis based on clinical data and medical images practices. These methods are Support Vector Machines (SVM), Decision Trees, Random Forest, Gradient Boosting, deep learning models, such as Convolutional Neural Networks (CNNs) and hybrid models among others. It has been argued that the ensemble and hybrid model can be used in not only high diagnostic accuracy of up to 98 but in part feature selection, dimensionality reduction, and data preprocessing on the model accuracy. Recent developments of contrastive learning, lightweight neural networks and explainable artificial intelligence (XAI) are addressed in order to make it more efficient and interpretable. Although these positive results are achieved, the issues that still exist are: the diversity of the datasets, the absence of the multi-center validation, and the complexity of the calculations. The results indicate that machine learning-based systems can be used as viable and scalable decision support systems in detecting breast cancer in its early stages.