Breast Cancer Detection through Histological Imaging: A Machine Learning Approach

Breast cancer is one of the most common types of cancer globally, and early detection is crucial for increasing the chances of survival. Mammography and biopsy are conventional diagnostic methods that are accurate but labor-intensive and prone to human error. Recent machine learning (ML)-based advancements have enabled automated systems for cancer classification that may improve their efficiency. In this study, using histological images of size 700 × 460, a total of 1,148 images, the performance of K-Nearest Neighbor (KNN) classification algorithm for breast cancer classification was evaluated. We split the data, where 70% is trained and 30% is tested. To enhance classification accuracy, various data preprocessing methods and feature selection techniques are implemented. The Results show that KNN is offering another fine Performance with Accuracy, Precision, Recall, and F1 score of 100% as the perfect prediction. Choose one optimal k value such that it provides best classification between (benign and malignant cases), which make the KNN one of the most accurate algorithms for breast cancer classification. The study signifies a paradigm shift in medical image analysis, indicating the efficacy of ML-based approaches over traditional approaches.