Application of graph-based features in computer-aided diagnosis for histopathological image classification of gastric cancer

Background: The gold standard for gastric cancer detection is histopathological image analysis, but there are certain drawbacks in the existing and diagnosis. Method: In this paper, based on study of computer-aided diagnosis (CAD) system, graph-based features applied to histopathology microscopic a classifier used classify cells from benign cells. Firstly, segmentation performed. After finding region, cell nuclei extracted using k-means method, minimum spanning tree (MST) drawn, MST extracted. then put into classification. Result: Different methods compared tissue stage, among which Level-Set, Otsu thresholding, watershed, SegNet, U-Net Trans-U-Net segmentation; Graph-based features, Red, Green, Blue Grey-Level Co-occurrence Matrix Histograms Oriented Gradient Local Binary Patterns feature extraction stage; Radial Basis Function (RBF) Support Vector Machine (SVM), Linear SVM, Artificial Neural Network, Random Forests, k-NearestNeighbor, VGG16, Inception-V3 stage. It found that segment areas, extracting finally RBF SVM gives optimal results with 94.29%. Conclusion: This paper focus analysis method histopathology. final experimental data shows our better than other classifying images cancer.