基于人工鱼群的淋巴恶性肿瘤细胞分割方法

Segmentation Algorithm of Lymphatic Malignant Tumor Cell Based on Artificial Fish-Swarm

研究淋巴恶性肿瘤细胞分割的问题,提高淋巴恶性肿瘤多细胞分割的准确率。针对淋巴恶性肿瘤细胞聚合度比较大,传统方式进行细胞分割时,很难根据病变细胞之间的灰度值差异分割聚合在一起的细胞。但是由于淋巴深度癌变的聚合细胞灰度值差异比较小,导致传统方式进行细胞分割的准确率比较低。为了避免上述问题,提出了一种人工鱼群算法的淋巴恶性肿瘤细胞分割方法,根据建立淋巴恶性肿瘤细胞轮廓模型,获取的淋巴恶性肿瘤细胞边界,利用人工鱼群边界聚类方法,对细胞进行分割处理。实验证明,提出的细胞分割方法能够提高淋巴恶性肿瘤细胞分割的准确率,取得了很好的效果。

We have researched the lymphatic malignant tumor cell division, and improved accuracy of lymphatic malignant tumor cell division. For the polymerization degree of lymphatic malignant tumor cell is higher and the grey value difference of lymphatic depth cancerous is smaller, a traditional way of cell division is difficult to divide those cells according to the differences of grey values of pathological change ceils, which results in low accuracy of segmen- tation. In order to avoid this problem, this paper proposed segmentation algorithm of lymphatic malignant tumor cells based on artificial fish. According to the established lymphatic malignant tumor cell contour model, we obtained the boundary of lymphatic malignant tumor cells, and used artificial fish boundary clustering method to divide the cells. Experiments show that the cell division algorithm can improve the accuracy of lymphatic malignant tumor cell divi-sion, and achieve good results.