Segmentation of pulmonary nodules in CT images based on 3D-UNET combined with three-dimensional conditional random field optimization.

PURPOSE Pulmonary nodules are a potential manifestation of lung cancer. In computer-aided diagnosis (CAD) of lung cancer, it is of great significance to extract the complete boundary of the pulmonary nodules in the CT (computed tomography) scans accurately. It can provide doctors with important information such as tumor size and density, which assist doctors in subsequent diagnosis and treatment. In addition to this, in the molecular subtype and radiomics of lung cancer, segmentation of lung nodules also plays a pivotal role. Existing methods are difficult to use only one model to simultaneously treat the boundaries of multiple types of lung nodules in CT images. METHOD In order to solve the problem, this paper proposed a 3D-UNET network model optimized by 3D-CRF (three-dimensional conditional random field) to segment pulmonary nodules. On the basis of 3D-UNET, the 3D-CRF is used to optimize the sample output of the training set, so as to update the network weights in training process, reduce the model training time and reduce the loss rate of the model. We selected 936 sets of pulmonary nodules data onto the Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) 1 database to train and test the model. What's more, we used clinical data from partner hospitals for additional validation. RESULTS AND CONCLUSIONS The results show that our method is accurate and effective. Particularly, it shows more significant for the optimization of the segmentation of adhesive pulmonary nodules (the juxta-pleural and juxta-vascular nodules) and ground glass pulmonary nodules (GGNs).

[1]  Zhaoxuan Gong,et al.  Pulmonary nodules segmentation method based on auto-encoder , 2018, International Conference on Digital Image Processing.

[2]  Lianfang Tian,et al.  Segmentation of ground glass opacity pulmonary nodules using an integrated active contour model with wavelet energy-based adaptive local energy and posterior probability-based speed function , 2016 .

[3]  Xiaofei Yan,et al.  Juxta-Vascular Pulmonary Nodule Segmentation in PET-CT Imaging Based on an LBF Active Contour Model with Information Entropy and Joint Vector , 2018, Comput. Math. Methods Medicine.

[4]  Anthony P. Reeves,et al.  Three-dimensional segmentation and growth-rate estimation of small pulmonary nodules in helical CT images , 2003, IEEE Transactions on Medical Imaging.

[5]  Xiaodong Wu,et al.  3D fully convolutional networks for co-segmentation of tumors on PET-CT images , 2018, 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018).

[6]  Richard C. Pais,et al.  The Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI): a completed reference database of lung nodules on CT scans. , 2011, Medical physics.

[7]  Bram van Ginneken,et al.  A large-scale evaluation of automatic pulmonary nodule detection in chest CT using local image features and k-nearest-neighbour classification , 2009, Medical Image Anal..

[8]  Kunio Doi,et al.  Computer-aided diagnostic scheme for distinction between benign and malignant nodules in thoracic low-dose CT by use of massive training artificial neural network , 2005, IEEE Transactions on Medical Imaging.

[9]  Elsa D. Angelini,et al.  Discriminative Localization in CNNs for Weakly-Supervised Segmentation of Pulmonary Nodules , 2017, MICCAI.

[10]  Hoon Ko,et al.  Automatic Lung Segmentation With Juxta-Pleural Nodule Identification Using Active Contour Model and Bayesian Approach , 2018, IEEE Journal of Translational Engineering in Health and Medicine.

[11]  Jiwen Lu,et al.  Left Atrial Appendage Segmentation Using Fully Convolutional Neural Networks and Modified Three-Dimensional Conditional Random Fields , 2018, IEEE Journal of Biomedical and Health Informatics.

[12]  Yong Li,et al.  Improved U-NET network for pulmonary nodules segmentation , 2018, Optik.

[13]  Luming Zhang,et al.  A mix-pooling CNN architecture with FCRF for brain tumor segmentation , 2019, J. Vis. Commun. Image Represent..

[14]  Richard C. Pais,et al.  Evaluation of Lung MDCT Nodule Annotation Across Radiologists and Methods 1 , 2006 .

[15]  Lubomir M. Hadjiiski,et al.  Computer-aided diagnosis of pulmonary nodules on CT scans: segmentation and classification using 3D active contours. , 2006, Medical physics.

[16]  Jin Mo Goo,et al.  Ground-glass nodule segmentation in chest CT images using asymmetric multi-phase deformable model and pulmonary vessel removal , 2018, Comput. Biol. Medicine.

[17]  Zhenyu Liu,et al.  Central focused convolutional neural networks: Developing a data-driven model for lung nodule segmentation , 2017, Medical Image Anal..

[18]  Seyed-Ahmad Ahmadi,et al.  Automatic Liver and Lesion Segmentation in CT Using Cascaded Fully Convolutional Neural Networks and 3D Conditional Random Fields , 2016, MICCAI.

[19]  Caiming Zhang,et al.  A fast weak-supervised pulmonary nodule segmentation method based on modified self-adaptive FCM algorithm , 2017, Soft Computing.

[20]  J. Austin,et al.  Guidelines for management of small pulmonary nodules detected on CT scans: a statement from the Fleischner Society. , 2005, Radiology.

[21]  Jianwei Wang,et al.  Joint learning for pulmonary nodule segmentation, attributes and malignancy prediction , 2018, 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018).

[22]  Xiaolin Huang,et al.  Pulmonary nodule segmentation with CT sample synthesis using adversarial networks , 2019, Medical physics.

[23]  Peipei Wang,et al.  Brain Tumor Segmentation Based on 3D Unet with Multi-Class Focal Loss , 2018, 2018 11th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI).

[24]  Aly A. Farag,et al.  Data-Driven Lung Nodule Models for Robust Nodule Detection in Chest CT , 2010, 2010 20th International Conference on Pattern Recognition.