Multi-scale input mirror network for tuberculosis detection in CXR image

Abstract

Computer-aided diagnosis (CAD) can detect tuberculosis (TB) cases, providing radiologists with more accurate and efficient diagnostic solutions. Various noise information in TB chest X-ray (CXR) images is a major challenge in this classification task. This study aims to propose a model with high performance in TB CXR image detection named multi-scale input mirror network(MIM-Net) based on CXR image symmetry, which consists of a multi-scale input feature extraction network and mirror loss. The multi-scale image input can enhance feature extraction, while the mirror loss can improve the network performance through self-supervision. We used a publicly available TB CXR image classification dataset to evaluate our proposed method via 5-fold cross-validation, with accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and area under curve (AUC) of 99.67%, 100%, 99.60%, 99.80%, 100%, and 0.999 9, respectively. Compared to other models, MIM-Net performed best in all metrics. Therefore, the proposed MIM-Net can effectively help the network learn more features and can be used to detect TB in CXR images, thus assisting doctors in diagnosing.

Keywords

computer-aided diagnosis (CAD) / medical image classification / deep learning / feature symmetry / mirror loss

References

[1] TASCI E. Pre-processing effects of the tuberculosis chest X-ray images on pre-trained CNNs: an investigation//Artificial Intelligence and Applied Mathematics in Engineering Problems, April 20-22, 2019, Antalya, Turkey. New York: Springer International Publishing, 2020: 589-596.

[2] CHATTOPADHYAY S, KUNDU R, SINGH P K, et al. Pneumonia detection from lung X-ray images using local search aided sine cosine algorithm based deep feature selection method. International Journal of Intelligent Systems, 2022, 37(7): 3777-3814.

[3] ZENG J B, LIU Z, SHEN G L, et al. MRI evaluation of pulmonary lesions and lung tissue changes induced by tuberculosis. International Journal of Infectious Diseases, 2019, 82: 138-146.

[4] JAIN S K, ANDRONIKOU S, GOUSSARD P, et al. Advanced imaging tools for childhood tuberculosis: potential applications and research needs. The Lancet Infectious Diseases, 2020, 20(11): e289-e297.

[5] DHOOT R, HUMPHREY J M, O'MEARA P, et al. Implementing a mobile diagnostic unit to increase access to imaging and laboratory services in western Kenya. BMJ Global Health, 2018, 3(5): e000947.

[6] LANGE C, MORI T. Advances in the diagnosis of tuberculosis. Respirology, 2010, 15(2): 220-240.

[7] QIN C L, YAO D M, SHI Y H, et al. Computer-aided detection in chest radiography based on artificial intelligence: a survey. Biomedical Engineering Online, 2018, 17(1): 113.

[8] KIM T K, YI P H, WEI J C, et al. Deep learning method for automated classification of anteroposterior and posteroanterior chest radiographs. Journal of Digital Imaging, 2019, 32(6): 925-930.

[9] LI X K, ZHOU Y K, DU P, et al. A deep learning system that generates quantitative CT reports for diagnosing pulmonary Tuberculosis. Applied Intelligence, 2021, 51(6): 4082-4093.

[10] LAKHANI P, SUNDARAM B. Deep learning at chest radiography: automated classification of pulmonary tuberculosis by using convolutional neural networks. Radiology, 2017, 284(2): 574-582.

[11] YI P H, KIM T K, WEI J C, et al. Automated semantic labeling of pediatric musculoskeletal radiographs using deep learning. Pediatric Radiology, 2019, 49(8): 1066-1070.

[12] KANT S, SRIVASTAVA M M. Towards automated tuberculosis detection using deep learning//2018 IEEE Symposium Series on Computational Intelligence, November 18-21, 2018, Bangalore, India. New York: IEEE, 2018: 1250-1253.

[13] SANTOSH K C, ANTANI S. Automated chest X-ray screening: can lung region symmetry help detect pulmonary abnormalities?IEEE Transactions on Medical Imaging, 2018, 37(5): 1168-1177.

[14] CHANDRA T B, VERMA K, SINGH B K, et al. Automatic detection of tuberculosis related abnormalities in Chest X-ray images using hierarchical feature extraction scheme. Expert Systems with Applications, 2020, 158: 113514. New York: IEEE, 2021: 408-413.

[15] SONI A, RAI A, AHIRWAR S K. Mycobacterium tuberculosis detection using support vector machine classification approach//2021 10th IEEE International Conference on Communication Systems and Network Technologies, June 18-19, 2021, Bhopal, India. New York: IEEE, 2021: 408-413.

[16] LI L J, HUANG H Y, JIN X Y. AE-CNN classification of pulmonary tuberculosis based on CT images//2018 9th International Conference on Information Technology in Medicine and Education, October 19-21, 2018, Hangzhou, China. New York: IEEE, 2018: 39-42.

[17] AHSAN M, GOMES R, DENTON A. Application of a Convolutional Neural Network using transfer learning for tuberculosis detection//2019 IEEE International Conference on Electro Information Technology, May 20-22, 2019. Brookings, SD, USA. New York: IEEE, 2019: 427-433.

[18] HUANG C X, WANG W, ZHANG X, et al. Tuberculosis diagnosis using deep transferred EfficientNet. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2023, 20(5): 2639-2646.

[19] MAHBUB M K, BISWAS M, GAUR L, et al. Deep features to detect pulmonary abnormalities in chest X-rays due to infectious diseaseX: Covid-19, pneumonia, and tuberculosis. Information Sciences, 2022, 592: 389-401.

[20] YANG J, JIN Y X, LIU Y B, et al. Research on fabric material recognition method based on improved transformer. Journal of North University of China (Natural Science Edition), 2023, 44(2): 138-145.

[21] DUONG L T, LE N H, TRAN T B, et al. Detection of tuberculosis from chest X-ray images: Boosting the performance with vision transformer and transfer learning. Expert Systems with Applications, 2021, 184: 115519.

[22] OKOLO G I, KATSIGIANNIS S, RAMZAN N. IEViT: An enhanced vision transformer architecture for chest X-ray image classification. Computer Methods and Programs in Biomedicine, 2022, 226: 107141.

[23] IOFFE S, SZEGEDY C, PARANHOS L, et al. Batch normalization: accelerating deep network training by reducing internal covariate shift. 2015: 1502.03167.

[24] MURPHY K, SCHÖLKOPF B, SRIVASTAVA N, et al. Dropout: a simple way to prevent neural networks from overfitting. Journal of Machine Learning Research, 2014, 15(1): 1929-1958.

[25] RAHMAN T, KHANDAKAR A, KADIR M A, et al. Reliable tuberculosis detection using chest X-ray with deep learning, segmentation and visualization. IEEE Access, 2020, 8: 191586-191601.

[26] KINGMA D P, BA J. Adam: A method for stochastic optimization. ArXiv preprint:1412.6980, 2014.

[27] LOSHCHILOV I, HUTTER F. Decoupled weight decay regularization. ArXiv preprint: 1711.05101, 2017.

[28] LOSHCHILOV I, HUTTER F. Decoupled weight decay regularization. ArXiv e-Prints, 2017: arXiv: 1711.05101.

[29] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks. Communications of the ACM, 2017, 60(6): 84-90.

[30] SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition. 2014: 1409.1556.

[31] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition//2016 IEEE Conference on Computer Vision and Pattern Recognition, June 27-30, 2016, Las Vegas, NV, USA. New York: IEEE, 2016: 770-778.

[32] HUANG G, LIU Z, VAN DER MAATEN L, et al. Densely connected convolutional networks//2017 IEEE Conference on Computer Vision and Pattern Recognition, July 21-26, 2017, Honolulu, HI, USA. New York: IEEE, 2017: 2261-2269.

[33] HU J, SHEN L, SUN G. Squeeze-and-excitation networks//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.

[34] SELVARAJU R R, COGSWELL M, DAS A, et al. Grad-CAM: visual explanations from deep networks via gradient-based localization. International Journal of Computer Vision, 2020, 128(2): 336-359.