Segmentation-guided network for automatic thoracic pathology classification
Tóm tắt
Lung diseases are the top causes of death around the world nowadays. Chest X-rays (CXRs) provide an invaluable tool for diagnosing lung-related diseases at the earliest stage possible. However, the accuracy of the diagnosis results depends heavily on the skill of the radiologist and is inevitably time-consuming and subjective. Accordingly, the present study proposes a model-based learning approach for the automatic detection of thoracic disease from CXR images designated as Segmentation-Guided Thorax Classification (SGTC). The proposed method consists of two stages, namely lung segmentation and thorax classification. The lung segmentation stage applies the U-Net model with ResNet-50 as the backbone to segment the lung region in the CXR. The thorax classification stage then utilizes the ChexNet model with DenseNet-121 as the backbone to evaluate the probability of 14 different thoracic pathologies. The experimental results show that SGTC achieves an average area under the receiver operating characteristic (ROC) curve (AUC) of 0.844 when applied to the ChestX-ray14 dataset. The performance of the proposed method is comparable to that of other recent approaches. Moreover, SGTC additionally superimposes a localization heatmap on the CXR image, which further assists the radiologist in interpreting the image.
Tài liệu tham khảo
Barnes PJ, Blasi F, Ward B, Reeves E, Rabe KF. Respiratory diseases in the world: realities of today - opportunities for tomorrow. Eur Respir J. 2014;42(3):3–5.
Bradley AP. The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognit. 1997;30(7):1145–59. https://doi.org/10.1016/S0031-3203(96)00142-2.
Candemir S, Jaeger S, Palaniappan K, Musco JP, et al. Lung segmentation in chest radiographs using anatomical atlases with nonrigid registration. IEEE Trans Med Imaging. 2014;33(2):577–90. https://doi.org/10.1109/TMI.2013.2290491.
Deng J, Dong W, Socher R, Li LJ, Li K, and Li FF. Imagenet: a large-scale hierarchical image database. in Proc IEEE Conf Comput Vis Pattern Recognit (CVPR). 2009;248–255. https://doi.org/10.1109/CVPR.2009.5206848.
Dice LR. Measures of the amount of ecologic association between species. Ecology. 1945;26(3):297–302. https://doi.org/10.2307/1932409.
Fawcett T. An introduction to ROC analysis. Pattern Recognit Lett. 2006;27(8):861–74. https://doi.org/10.1016/j.patrec.2005.10.010.
Ferkol T, Schraufnagel D. The global burden of respiratory disease. Ann Am Thorac Soc. 2014;11:404–6.
Ginneken BV, Stegmann MB, Loog M. Segmentation of anatomical structures in chest radiographs using supervised methods: a comparative study on a public database. Med Image Anal. 2006;10(1):19–40. https://doi.org/10.1016/j.media.2005.02.002.
Guan Q, Huang Y, Zhong Z, Zheng Z, Zheng L, Yang Y. Diagnose like a radiologist: attention guided convolutional neural network for thorax disease classification. arXiv. 2018;1801.09927. Retrieved January 20, 2020, from https://arxiv.org/abs/1801.09927.
He K, Zhang X, Ren S, and Sun J. Deep residual learning for image recognition. in Proc IEEE Conf Comput Vis Pattern Recognit (CVPR). 2016;770–778. https://doi.org/10.1109/CVPR.2016.90.
Hochreiter S, Schmidhuber J. Long short-term memory. Neural Comput. 1997;9(8):1735–80. https://doi.org/10.1162/neco.1997.9.8.1735.
Huang G, Liu Z, Maaten LVD, and Weinberger KQ. Densely connected convolutional networks. in Proc. 2017 IEEE Conf Comput Vis Pattern Recognit (CVPR). 2017;2261–2269. https://doi.org/10.1109/CVPR.2017.243.
Ioffe S and Szegedy C. Batch normalization: accelerating deep network training by reducing internal covariate shift. in Proc 32nd Int Conf Mach Learn (ICML). 2015;37:448–456.
Islam J, Zhang Y. Towards robust lung segmentation in chest radiographs with deep learning. In: Proc Mach Learn Health Workshop at NeurIPS2018. 2018. Retrieved December 11, 2019, from https://arxiv.org/abs/1811.12638.
Islam MT, Aowal MA, Minhaz AT, Ashraf K. Abnormality detection and localization in chest x-rays using deep convolutional neural networks. arXiv. 2017;1705.09850. Retrieved December 02, 2019, from https://arxiv.org/abs/1705.09850.
Jaeger S, Karargyris A, Candemir S, Folio L, et al. Automatic tuberculosis screening using chest radiographs. IEEE Trans Med Imaging. 2014;33(2):233–45. https://doi.org/10.1109/TMI.2013.2284099.
Kingma DP, Ba J. Adam: a method for stochastic optimization. In: Proc 3rd Int Conf Learn Represent (ICLR). 2015. Retrieved October 08, 2019, from https://arxiv.org/abs/1412.6980.
Kipf TN, Welling M. Semi-supervised classification with graph convolutional networks. In: Proc. Int. Conf Learn Represent (ICLR). 2017. Retrieved from May 12, 2020, from https://arxiv.org/abs/1609.02907.
Le HT, Pham HTT. Brain tumor segmentation using U-net based fully convolutional networks and extremely randomized trees. Vietnam J Sci Technol Eng. 2018;60(3):19–25. https://doi.org/10.31276/VJSTE.60(3).19.
Lowe DG. Distinctive image features from scale-invariant keypoints. Int J Comput Vis. 2004;60(2):91–110. https://doi.org/10.1023/B:VISI.0000029664.99615.94.
Mao C, Yao L, Luo Y, Mao C. ImageGCN: Multi-relational image graph convolutional networks for disease identification with chest x-rays. arXiv. 2019;1904.00325. Retrieved February 27, 2020, from https://arxiv.org/abs/1904.00325.
Nair V and Hinton GE. Rectified linear units improve restricted boltzmann machines. in Proc. the 27th Int. Conf Mach Learn (ICML). 2010;807–814.
Powers DMW. Evaluation: from precision, recall and F-factor to ROC, informedness, markedness and correlation. J Mach Learn Tech. 2011;2(1):37–63.
Rajpurkar P, Irvin J, Zhu K, Yang B, Mehta H, Duan T, et al. CheXNet: radiologist-level pneumonia detection on chest x-rays with deep learning. arXiv. 2017;1711.05225. Retrieved November 05, 2019, from https://arxiv.org/abs/1711.05225.
Ronneberger O, Fischer P, Brox T. U-Net: Convolutional networks for biomedical image segmentation. In: Proc. 18th Int Conf Med Image Comput Comput Assist Interv (MICCAI). 2015;9351:234–241. https://doi.org/10.1007/978-3-319-24574-4_28.
Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D, and Batra D. Grad-CAM: Visual explanations from deep networks via gradient-based localization. in Proc 2017 IEEE Int Conf Comput Vis (ICCV). 2017;618–626. https://doi.org/10.1109/ICCV.2017.74.
Shiraishi J, Katsuragawa S, Ikezoe J, Matsumoto T, et al. Development of a digital image database for chest radiographs with and without a lung nodule: receiver operating characteristic analysis of radiologists’ detection of pulmonary nodules. Am J Roentgenol. 2000;174(1):71–4. https://doi.org/10.2214/ajr.174.1.1740071.
Smith LN. Cyclical learning rates for training neural networks. in Proc IEEE Winter Conf App Comput Vis (WACV). 2017;464–472. https://doi.org/10.1109/WACV.2017.58.
Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R. Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res. 2014;15(56):1929–58.
Wang X, Peng Y, Lu L, Lu Z, Bagheri M, and Summers RM. ChestX-ray8: Hospital-scale chest x-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. in Proc. IEEE Conf Comp Vis Pattern Recognit. 2017;2097–2106. https://doi.org/10.1109/CVPR.2017.369.
Wang CM, Elazab A, Jia FC, Wu JH, Wu QM. Automated chest screening based on hybrid model of transfer learning and convolutional sparse denoising autoencorder. Biomed Eng Online. 2018;17:63. https://doi.org/10.1186/s12938-018-0496-2.
Yao L, Poblenz E, Dagunts D, Covington B, Bernard D, Lyman K. Learning to diagnose from scratch by exploiting dependencies among labels. arXiv. 2018;1710.10501v2. Retrieved March 12, 2020, from https://arxiv.org/abs/1710.10501.
Zhou B, Khosla A, Lapedriza A, Oliva A, and Torralba A. Learning deep features for discriminative localization. in Proc IEEE Conf Comp Vis Pattern Recognit (CVPR). 2016;2921–2929. https://doi.org/10.1109/CVPR.2016.319.