Lung Nodule Classification Algorithm Based on Fusion Features
Abstract— Lung nodules are the lesion areas of lung, and malignant lung nodules may led to lung cancer. Nowadays, computer-aided diagnosis based on CT images are important for lung cancer. However, the existing methods of feature extraction of CT image did not perform effect. This paper proposes the DSS algorithm for lung nodules classification based on fusion features. The DSS adopts the local-global model to extract the depth features of images, and jointly uses the shape descriptors based on medical knowledge. Besides, it combines the fusion features into the Support Vector Machine (SVM) for lung nodule classification. This paper has evaluated the DSS algorithm on LIDC-IDRI data set and the method performs effect for lung nodules classification.
Index Terms— Lung nodule; CT image; DSS; fusion features; SVM
Software School, Xiamen University, CHINA
Cite: Shengyu Lu, "Lung Nodule Classification Algorithm Based on Fusion Features," Proceedings of 2019 the 9th International Workshop on Computer Science and Engineering, pp. 622-629, Hong Kong, 15-17 June, 2019.