多特征融合的乳腺癌组织病理学图像识别的方法

Breast cancer histopathological image recognition based on multi-feature fusion

乳腺癌是全球女性常见的癌症类型之一,严重影响了女性的健康,乳腺癌组织病理学图像的识别已成为医学图像处理领域的研究热点。针对Bioimaging 2015数据集进行乳腺癌组织病理学图像的识别研究,将该数据集分为癌类与非癌类2种。实验提取了乳腺癌组织病理学图像染色分离后4个方向上的灰度共生矩阵特征、小波特征及Tamura纹理特征,并根据颜色自动相关图提取了原始图像的颜色特征,同时也提取了染色分离前水平方向上的灰度共生矩阵特征作为纹理信息的补充,最后将提取到的特征进行融合,并输入到支持向量机分类器中,以实现乳腺癌组织病理学图像的识别,识别准确率达到了83.33%。

Breast cancer is one of the most common types of cancer in women around the world,and seriously affects the health of women.The recognition of histopathological images of breast cancer has become a research hotspot in the field of medical image processing.The Bioimaging 2015 dataset is used to recognize breast cancer histopathological images,and the dataset is classified into two types:cancer and non-cancer.In the experiments,the features of gray-level co-occurrence matrix in four directions,the wavelet features and Tamura texture features are extracted from the stained-separated breast cancer histopathological images.And the color features of the original images are extracted according to the color auto-correlogram.At the same time,the gray-level co-occurrence matrix features in the horizontal direction before stain separation are extracted as a supplement to the texture information.Finally,the extracted features are fused and input into the support vector machine classifier to realize the recognition of breast cancer histopathological images,the recognition accuracy is 83.33%.