论文信息 - Pulmonary Nodule Classification in Computed Tomography Image Using a 3D Deep Convolutional Neural Network

Pulmonary Nodule Classification in Computed Tomography Image Using a 3D Deep Convolutional Neural Network

Early detection and examination of pulmonary nodules is the most effective ways to prevent lung cancer, accounting for more than a quarter of all cancer deaths. In this paper, we propose a 3D deep convolutional neural network for pulmonary nodule recognition. We use deep convolutional neural network that uses shortcut connections and the ensemble method is used to boost recognition performance. Proposed models are trained and tested on Lung Nodule Analysis 2016 competition dataset. We evaluate performance of models and verify preciseness. Proposed model produces 0.899 of Competition Performance Metric value, that is evaluation criteria of competition. It is outperforming value than that of other participants.

[1] Geoffrey E. Hinton,et al. Learning representations by back-propagating errors , 1986, Nature.

[2] Mark J. van der Laan,et al. The relative performance of ensemble methods with deep convolutional neural networks for image classification , 2017, Journal of applied statistics.

[3] Yongdong Zhang,et al. Automated pulmonary nodule detection in CT images using deep convolutional neural networks , 2019, Pattern Recognit..

[4] Hao Chen,et al. Multilevel Contextual 3-D CNNs for False Positive Reduction in Pulmonary Nodule Detection , 2017, IEEE Transactions on Biomedical Engineering.

[5] Jian Sun,et al. Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[6] Scott Lundberg,et al. Checkpoint Ensembles: Ensemble Methods from a Single Training Process , 2017, ArXiv.

[7] Michael R Hamblin,et al. CA : A Cancer Journal for Clinicians , 2011 .

[8] Adin Ramirez Rivera,et al. Large residual multiple view 3D CNN for false positive reduction in pulmonary nodule detection , 2017, 2017 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB).