基于三维卷积神经网络的肺结节识别研究
|
摘要
针对传统计算机辅助检测系统中肺结节检测存在大量假阳性的问题,提出一种基于三维卷积神经网络的肺结节识别方法。首先,将传统二维卷积神经网络扩展为三维卷积神经网络,充分挖掘肺结节的三维特征,增强特征的表达能力;其次,将密集连接网络与SENet相结合,在加强特征传递和复用的同时,通过特征重标定自适应学习特征权重;另外,引入focal loss作为网络的分类损失函数,提高对难样本的学习。在LUNA16数据集上的实验结果表明:与当前的主流深度学习算法相比,所提网络模型在平均每组CT图像中假阳个数为1和4时的检出率达到了0.911和0.934,CPM得分为0.891,优于大部分主流算法。
Herein, a method of pulmonary nodule recognition based on a three-dimensional(3 D) convolution neural network(CNN) is proposed to overcome the problem of false positives in pulmonary nodule detection by traditional computer aided detection systems. First, a traditional two-dimensional CNN is extended to 3 D CNN to fully extract the 3 D features of pulmonary nodules and enhance the expressive ability of the features. Second, dense connection network and SENet are combined to enhance feature transfer and reuse, and feature weights are adaptively learned by feature recalibration. In addition, focal loss is introduced as the network classification loss to improve the learning of hard examples. The experimental results on the LUNA16 dataset demonstrate that the proposed network model achieves sensitivities of 0.911 and 0.934 at one and four false positives per scan, respectively, and the competition performance metric is up to 0.891, which is better than that of existing mainstream methods.
Herein, a method of pulmonary nodule recognition based on a three-dimensional(3 D) convolution neural network(CNN) is proposed to overcome the problem of false positives in pulmonary nodule detection by traditional computer aided detection systems. First, a traditional two-dimensional CNN is extended to 3 D CNN to fully extract the 3 D features of pulmonary nodules and enhance the expressive ability of the features. Second, dense connection network and SENet are combined to enhance feature transfer and reuse, and feature weights are adaptively learned by feature recalibration. In addition, focal loss is introduced as the network classification loss to improve the learning of hard examples. The experimental results on the LUNA16 dataset demonstrate that the proposed network model achieves sensitivities of 0.911 and 0.934 at one and four false positives per scan, respectively, and the competition performance metric is up to 0.891, which is better than that of existing mainstream methods.
引文
[1] Sluimer I,Schilham A,Prokop M,et al.Computer analysis of computed tomography scans of the lung:a survey[J].IEEE Transactions on Medical Imaging,2006,25(4):385-405.
[2] Wang C M.Key techniques for lung nodule detection and classification based on chest imaging[D].Shenzhen:University of Chinese Academy of Sciences,2018:1-3.王昌淼.基于胸部影像的肺结节检测与分类关键技术研究[D].深圳:中国科学院大学,2018:1-3.
[3] Messay T,Hardie R C,Rogers S K.A new computationally efficient CAD system for pulmonary nodule detection in CT imagery[J].Medical Image Analysis,2010,14(3):390-406.
[4] Setio A A A,Jacobs C,Gelderblom J,et al.Automatic detection of large pulmonary solid nodules in thoracic CT images[J].Medical Physics,2015,42(10):5642-5653.
[5] Setio A A A,Ciompi F,Litjens G,et al.Pulmonary nodule detection in CT images:false positive reduction using multi-view convolutional networks[J].IEEE Transactions on Medical Imaging,2016,35(5):1160-1169.
[6] Dou Q,Chen H,Yu L Q,et al.Multilevel contextual 3-D CNNs for false positive reduction in pulmonary nodule detection[J].IEEE Transactions on Biomedical Engineering,2017,64(7):1558-1567.
[7] Jin H S,Li Z Y,Tong R F,et al.A deep 3D residual CNN for false-positive reduction in pulmonary nodule detection[J].Medical Physics,2018,45(5):2097-2107.
[8] Dou Q,Chen H,Jin Y M,et al.Automated pulmonary nodule detection via 3D ConvNets with online sample filtering and hybrid-loss residual learning[M]∥Descoteaux M,Maier-Hein L,Franz A,et al.Springer Proceedings in Physics.Singapore:Springer,2017:630-638.
[9] Huang G,Liu Z,Maaten L V D,et al.Densely connected convolutional networks[C]∥2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR),July 21-26,2017,Honolulu,HI,USA.New York:IEEE,2017:2261-2269.
[10] Hu J,Shen L,Sun G.Squeeze-and-excitation networks[C]∥2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition,June 18-23,2018,Salt Lake City,UT,USA.New York:IEEE,2018:7132-7141.
[11] Lin T Y,Goyal P,Girshick R,et al.Focal loss for dense object detection[C]//2017 IEEE International Conference on Computer Vision (ICCV),October 22-29,2017,Venice,Italy.New York:IEEE,2017:2999-3007.
[12] Li Z Q,Zhu R F,Gao F,et al.Hyperspectral remote sensing image classification based on three-dimensional convolution neural network combined with conditional random field optimization[J].Acta Optica Sinica,2018,38(8):0828001.李竺强,朱瑞飞,高放,等.三维卷积神经网络模型联合条件随机场优化的高光谱遥感影像分类[J].光学学报,2018,38(8):0828001.
[13] Lü X Q,Wu L,Gu Y,et al.Detection of low dose CT pulmonary nodules based on 3D convolution neural network[J].Optics and Precision Engineering,2018,26(5):1211-1218.吕晓琪,吴凉,谷宇,等.基于三维卷积神经网络的低剂量CT肺结节检测[J].光学精密工程,2018,26(5):1211-1218.
[14] Miao G,Li C F.Detection of pulmonary nodules CT images combined with two-dimensional and three-dimensional convolution neural networks[J].Laser & Optoelectronics Progress,2018,55(5):051006.苗光,李朝锋.二维和三维卷积神经网络相结合的CT图像肺结节检测方法[J].激光与光电子学进展,2018,55(5):051006.
[15] Setio A A A,Traverso A,de Bel T,et al.Validation,comparison,and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images:the LUNA16 challenge[J].Medical Image Analysis,2017,42:1-13.
[16] Messay T,Hardie R C,Tuinstra T R.Segmentation of pulmonary nodules in computed tomography using aregression neural network approach and its application to the lung image database consortium and image database resource initiative dataset[J].Medical Image Analysis,2015,22(1):48-62.
[17] He K M,Zhang X Y,Ren S Q,et al.Delving deep into rectifiers:surpassing human-level performance on ImageNet classification[C]//2015 IEEE International Conference on Computer Vision (ICCV),December 7-13,2015,Santiago,Chile.New York:IEEE,2015:1026-1034.
[18] Sakamoto M,Nakano H,Zhao K,et al.Multi-stage neural networks with single-sided classifiers for false positive reduction and its evaluation using lung X-Ray CT images[M]∥Battiato S,Gallo G,Schettini R,et al.Image Analysis and Processing-ICIAP 2017.Cham:Springer,2017:370-379.
[2] Wang C M.Key techniques for lung nodule detection and classification based on chest imaging[D].Shenzhen:University of Chinese Academy of Sciences,2018:1-3.王昌淼.基于胸部影像的肺结节检测与分类关键技术研究[D].深圳:中国科学院大学,2018:1-3.
[3] Messay T,Hardie R C,Rogers S K.A new computationally efficient CAD system for pulmonary nodule detection in CT imagery[J].Medical Image Analysis,2010,14(3):390-406.
[4] Setio A A A,Jacobs C,Gelderblom J,et al.Automatic detection of large pulmonary solid nodules in thoracic CT images[J].Medical Physics,2015,42(10):5642-5653.
[5] Setio A A A,Ciompi F,Litjens G,et al.Pulmonary nodule detection in CT images:false positive reduction using multi-view convolutional networks[J].IEEE Transactions on Medical Imaging,2016,35(5):1160-1169.
[6] Dou Q,Chen H,Yu L Q,et al.Multilevel contextual 3-D CNNs for false positive reduction in pulmonary nodule detection[J].IEEE Transactions on Biomedical Engineering,2017,64(7):1558-1567.
[7] Jin H S,Li Z Y,Tong R F,et al.A deep 3D residual CNN for false-positive reduction in pulmonary nodule detection[J].Medical Physics,2018,45(5):2097-2107.
[8] Dou Q,Chen H,Jin Y M,et al.Automated pulmonary nodule detection via 3D ConvNets with online sample filtering and hybrid-loss residual learning[M]∥Descoteaux M,Maier-Hein L,Franz A,et al.Springer Proceedings in Physics.Singapore:Springer,2017:630-638.
[9] Huang G,Liu Z,Maaten L V D,et al.Densely connected convolutional networks[C]∥2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR),July 21-26,2017,Honolulu,HI,USA.New York:IEEE,2017:2261-2269.
[10] Hu J,Shen L,Sun G.Squeeze-and-excitation networks[C]∥2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition,June 18-23,2018,Salt Lake City,UT,USA.New York:IEEE,2018:7132-7141.
[11] Lin T Y,Goyal P,Girshick R,et al.Focal loss for dense object detection[C]//2017 IEEE International Conference on Computer Vision (ICCV),October 22-29,2017,Venice,Italy.New York:IEEE,2017:2999-3007.
[12] Li Z Q,Zhu R F,Gao F,et al.Hyperspectral remote sensing image classification based on three-dimensional convolution neural network combined with conditional random field optimization[J].Acta Optica Sinica,2018,38(8):0828001.李竺强,朱瑞飞,高放,等.三维卷积神经网络模型联合条件随机场优化的高光谱遥感影像分类[J].光学学报,2018,38(8):0828001.
[13] Lü X Q,Wu L,Gu Y,et al.Detection of low dose CT pulmonary nodules based on 3D convolution neural network[J].Optics and Precision Engineering,2018,26(5):1211-1218.吕晓琪,吴凉,谷宇,等.基于三维卷积神经网络的低剂量CT肺结节检测[J].光学精密工程,2018,26(5):1211-1218.
[14] Miao G,Li C F.Detection of pulmonary nodules CT images combined with two-dimensional and three-dimensional convolution neural networks[J].Laser & Optoelectronics Progress,2018,55(5):051006.苗光,李朝锋.二维和三维卷积神经网络相结合的CT图像肺结节检测方法[J].激光与光电子学进展,2018,55(5):051006.
[15] Setio A A A,Traverso A,de Bel T,et al.Validation,comparison,and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images:the LUNA16 challenge[J].Medical Image Analysis,2017,42:1-13.
[16] Messay T,Hardie R C,Tuinstra T R.Segmentation of pulmonary nodules in computed tomography using aregression neural network approach and its application to the lung image database consortium and image database resource initiative dataset[J].Medical Image Analysis,2015,22(1):48-62.
[17] He K M,Zhang X Y,Ren S Q,et al.Delving deep into rectifiers:surpassing human-level performance on ImageNet classification[C]//2015 IEEE International Conference on Computer Vision (ICCV),December 7-13,2015,Santiago,Chile.New York:IEEE,2015:1026-1034.
[18] Sakamoto M,Nakano H,Zhao K,et al.Multi-stage neural networks with single-sided classifiers for false positive reduction and its evaluation using lung X-Ray CT images[M]∥Battiato S,Gallo G,Schettini R,et al.Image Analysis and Processing-ICIAP 2017.Cham:Springer,2017:370-379.