STACKPath: A Stacked Classifier-Based ML Pathway for Histopathological Cancer Detection

Breast cancer remains one of the leading causes of cancer-related deaths among women worldwide. However, early detection, especially in the precancerous stages, significantly improves the chances of complete recovery. Existing medical solutions face equipment constraints, scarcity of skilled professionals, and restricted access to advanced diagnostic technologies in rural and low-income areas. Additionally, manual interpretation of vast histopathological images is time-consuming, prone to human error, and lacks consistency. This paper proposes an AI-driven computer-aided diagnosis (CAD) system that leverages machine learning techniques to enhance breast cancer detection from histopathological images. Our approach involves innovative image preprocessing, feature extraction, and classification methods to improve the accuracy and efficiency of cancer detection based on multi-class classification. The pipeline integrates adaptive contrast enhancement, optimised feature selection, and advanced neural networks to identify abnormal lesions and assess the likelihood of malignancy. By training our model on a diverse dataset, we ensure robustness and generalisability across various demographic groups. This work advances computational pathology and explores novel AI methodologies to optimize each stage of the detection process. Ultimately, this work aims to integrate AI-assisted tools into clinical workflows, supporting pathologists in making more informed and timely diagnostic decisions.