Research on Classification of Breast Pathology Images Based on Transfer Learning
To assist the benign and malignant diagnosis of pathological section of breast tissue,this study utilizes the BreakHis public dataset to carry out breast tissue pathology images classification.By considering the staining differences and insufficient quantity of data,the staining normalization and data augmentation preprocessing were respectively carried out.By transfer learning,the fine-tuned ResNet 50 model is used to classify the preprocessed data,which results in the highest classification ac-curacy is 99.60％.The obtained results are compared with existing research outcomes,and proved that the effectiveness of the proposed method in addressing breast tissue image classification.
breast pathology imagestaining normalizationdata enhancementimage classification
万方数据
维普
