Histopathology Image Classification for Breast Cancer Using Convolutional Neural Network
Xinfei Gu · 2024
Breast cancer detection presents considerable challenges in terms of diagnostic accuracy and efficiency, particularly when relying on traditional manual examination techniques. As medical imaging data grows in scale and complexity, traditional machine learning methods struggle due to their dependence on manual feature extraction and limitations in modeling nonlinear relationships. Computer vision, a key area of artificial intelligence, plays a crucial role in automating image recognition tasks, and its application in medical imaging is becoming increasingly prominent. In this study, we propose a deep learning framework based on convolutional neural networks (CNN) for breast cancer histopathology image classification, a critical task within computer vision. By automatically learning hierarchical features from raw image data, our model bypasses the need for manual feature engineering, achieving an impressive classification accuracy of 94.26% and an F1 score of 0.97. The robustness and generalization ability of the model are especially evident in its performance on complex patterns found in invasive ductal carcinoma (IDC) images. Specifically, the CNNbased approach effectively captures distinct visual features such as edge details, complex textures, and differential staining, which are essential for precise classification in medical image recognition.