Segmentation of Pulmonary CT Image by Using Convolutional Neural Network Based on Membership Function

摘要

The accurate segmentation of pulmonary CT images is of great significance to clinical computer-aided diagnosis and treatment. In order to avoid the explicit extraction of image features and improve the efficiency of image segmentation and reduce the influence of human factors on the segmentation results, this paper proposes a method of segmenting pulmonary CT image based on membership function convolution neural network (MFCNN). First, the method uses pulmonary CT image filtered by the Gaussian as the input data of the convolution neural network. Then, that uses the improved convolution neural network to achieve the initial segmentation of the image. Finally, the final segmentation result is obtained by setting the threshold based on this paper method. After experimental comparison, this paper demonstrates the feasibility and effectiveness of convolution neural network in the segmentation of pulmonary CT images.