Breast Cancer Immunohistochemical Classification Network Based on Fine-Tuned Transfer Learning
Jiangtao Hu · 2024
Breast cancer is the leading cause of cancer-related deaths among women globally. Early and accurate diagnosis of breast cancer is of paramount importance. In recent years, transfer learning has made significant breakthroughs in the field of machine learning, and its application in the classification of breast cancer Her2 immunohistochemical images has emerged as a new research area. In this study, we propose a transfer learning framework based on the Densenet161 network for classifying breast cancer immunohistochemical images with staining scores of 0, 1+, 2+, and 3+ using immunohistochemical methods. In this framework, we first establish a transfer learning architecture based on Densenet161. We utilized fine-tuning techniques to extract highly effective features from the Densenet161 architecture for classification. A publicly available dataset was employed for comparative experiments, revealing that the Densenet161 model achieved an average accuracy of 97.07%.