Abstract
Tuberculosis was discovered about 130 years ago and has remained a persistent threat and leading cause of death worldwide. Drug therapy is currently one of the effective treatments for pulmonary tuberculosis. Tuberculosis can be divided into drug-sensitive and drug-resistant tuberculosis in terms of its drug resistance. The timely identification of drug-resistant from drug-sensitive tuberculosis is vital for effective clinical treatment and improving the cure rate of patients. The most commonly used methods of drug-resistant detection are either expensive or time-consuming (up to several months). Therefore, an effective and affordable method for early drug-resistant screening is urgently required. Automatic diagnosis using computed Tomography is one of the possible approaches for this requirement. Although there have been some applications of deep learning-based algorithms on this task, open-source datasets are lacking to support clinical significance. In this paper, a novel method is proposed to distinguish drug-resistant tuberculosis from drug-sensitive tuberculosis. First, an annotated dataset is constructed based on the clinical pathological report, in which 101 drug-sensitive cases and 304 drug-resistant cases are collected and preprocessed. Then, we employ a three-dimensional residual network as the backbone and propose a lesion slice selection strategy to train the constructed model effectively. At last, we reimplement existing methods using the constructed dataset for comparison. Extensive experimental results demonstrate that the proposed method performs the best classification with a rate of 92.21% and an AUC value of 93.50%, significantly exceeding the current baseline results.
This work was supported by the National Natural Science Foundation of China under Grant 62106163 and by the National Natural Science Foundation of China under Grant 82100119 and the CAAI-Huawei MindSpore Open Fund.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Aghajanzadeh, E., Shomali, B., Aminshahidi, D., Ghassemi, N.: Classification of tuberculosis type on CT scans of lungs using a fusion of 2d and 3d deep convolutional neural networks. In: CLEF (Working Notes), pp. 1133–1144 (2021)
Ahmed, M.S., Sk, O., Jayatilake, M., Gonçalves, T., Rato, L.: Texture analysis from 3d model and individual slice extraction for tuberculosis MDR detection, type classification and severity scoring (2018)
Cha, J., et al.: Radiological findings of extensively drug-resistant pulmonary tuberculosis in non-aids adults: comparisons with findings of multidrug-resistant and drug-sensitive tuberculosis. Korean J. Radiol. 10(3), 207–216 (2009)
Chen, S., Guo, J., Wang, C., Xu, X., Yi, Z., Li, W.: DeepLNAnno: a web-based lung nodules annotating system for CT images. J. Med. Syst. 43(7), 1–9 (2019). https://doi.org/10.1007/s10916-019-1258-9
Cui, E.N., et al.: Radiomics model for distinguishing tuberculosis and lung cancer on computed tomography scans. World J. Clin. Cases 8(21), 5203 (2020)
Gao, X., Comley, R., Khan, M.H.M.: An enhanced deep learning architecture for classification of tuberculosis types from CT lung images. In: 2020 IEEE International Conference on Image Processing (ICIP), pp. 2486–2490. IEEE (2020)
Gao, X.W., James-Reynolds, C., Currie, E.: Analysis of tuberculosis severity levels from CT pulmonary images based on enhanced residual deep learning architecture. Neurocomputing 392, 233–244 (2020)
Gao, X.W., Qian, Y.: Prediction of multidrug-resistant TB from CT pulmonary images based on deep learning techniques. Mol. Pharm. 15(10), 4326–4335 (2017)
Hankare, S., Shirguppikar, S.: Detection of tuberculosis and lung cancer using CNN. In: Handbook of Smart Materials, Technologies, and Devices: Applications of Industry 4.0, pp. 1–11. Springer (2021)
Hara, K., Kataoka, H., Satoh, Y.: Can spatiotemporal 3d CNNS retrace the history of 2d cnns and imagenet? In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 6546–6555 (2018)
Janowczyk, A., Madabhushi, A.: Deep learning for digital pathology image analysis: a comprehensive tutorial with selected use cases. J. Pathol. Inf. 7(1), 29 (2016)
Ji, S., Xu, W., Yang, M., Yu, K.: 3d convolutional neural networks for human action recognition. IEEE Trans. Pattern Anal. Mach. Intell. 35(1), 221–231 (2012)
Kant, S., Srivastava, M.M.: Towards automated tuberculosis detection using deep learning. In: 2018 IEEE Symposium Series on Computational Intelligence (SSCI), pp. 1250–1253. IEEE (2018)
Kay, W., et al.: The kinetics human action video dataset. arXiv preprint. arXiv:1705.06950 (2017)
Kazlouski, S.: Tuberculosis CT image analysis using image features extracted by 3d autoencoder. In: Arampatzis, A. (ed.) CLEF 2020. LNCS, vol. 12260, pp. 131–140. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58219-7_12
Li, L., Huang, H., Jin, X.: AE-CNN classification of pulmonary tuberculosis based on CT images. In: 2018 9th International Conference on Information Technology in Medicine and Education (ITME), pp. 39–42. IEEE (2018)
Li, X., Zhou, Y., Du, P., Lang, G., Xu, M., Wu, W.: A deep learning system that generates quantitative CT reports for diagnosing pulmonary tuberculosis. Appl. Intell. 51(6), 4082–4093 (2021). https://doi.org/10.1007/s10489-020-02051-1
O’Brien, R.J.: Drug-resistant tuberculosis: etiology, management and prevention. In: Seminars in respiratory infections, vol. 9, pp. 104–112 (1994)
Rahman, T., et al.: Reliable tuberculosis detection using chest x-ray with deep learning, segmentation and visualization. IEEE Access 8, 191586–191601 (2020)
Selvaraju, R.R., Das, A., Vedantam, R., Cogswell, M., Parikh, D., Batra, D.: Grad-cam: why did you say that? arXiv preprint. arXiv:1611.07450 (2016)
Wang, X., Peng, Y., Lu, L., Lu, Z., Bagheri, M., Summers, R.M.: Chestx-ray8: hospital-scale chest x-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2097–2106 (2017)
WHO, G.: Global tuberculosis report 2020. Glob. Tuberc. Rep (2020)
Wong, A., Lee, J.R.H., Rahmat-Khah, H., Sabri, A., Alaref, A., Liu, H.: TB-Net: a tailored, self-attention deep convolutional neural network design for detection of tuberculosis cases from chest x-ray images. Front. Artif. Intell. 5 (2022)
Xie, Y., et al.: Knowledge-based collaborative deep learning for benign-malignant lung nodule classification on chest CT. IEEE Trans. Med. Imaging 38(4), 991–1004 (2018)
Xu, X., et al.: DeepLN: a framework for automatic lung nodule detection using multi-resolution CT screening images. Knowl.-Based Syst. 189, 105128 (2020)
Zhou, K., et al.: Automatic airway tree segmentation based on multi-scale context information. Int. J. Comput. Assist. Radiol. Surg. 16(2), 219–230 (2021). https://doi.org/10.1007/s11548-020-02293-x
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Du, Q. et al. (2022). Automatic Diagnose of Drug-Resistance Tuberculosis from CT Images Based on Deep Neural Networks. In: Fang, L., Povey, D., Zhai, G., Mei, T., Wang, R. (eds) Artificial Intelligence. CICAI 2022. Lecture Notes in Computer Science(), vol 13605. Springer, Cham. https://doi.org/10.1007/978-3-031-20500-2_21
Download citation
DOI: https://doi.org/10.1007/978-3-031-20500-2_21
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-20499-9
Online ISBN: 978-3-031-20500-2
eBook Packages: Computer ScienceComputer Science (R0)