Vulnerability in Deep Transfer Learning Models to Adversarial Fast Gradient Sign Attack for COVID-19 Prediction from Chest Radiography Images
Source abstract
Abstract
The COVID-19 pandemic requires the rapid isolation of infected patients. Thus, high-sensitivity radiology images could be a key technique to diagnose patients besides the polymerase chain reaction approach. Deep learning algorithms are proposed in several studies to detect COVID-19 symptoms due to the success in chest radiography image classification, cost efficiency, lack of expert radiologists, and the need for faster processing in the pandemic area. Most of the promising algorithms proposed in different studies are based on pre-trained deep learning models. Such open-source models and lack of variation in the radiology image-capturing environment make the diagnosis system vulnerable to adversarial attacks such as fast gradient sign method (FGSM) attack. This study therefore explored the potential vulnerability of pre-trained convolutional neural network algorithms to the FGSM attack in terms of two frequently used models, VGG16 and Inception-v3. Firstly, we developed two transfer learning models for X-ray and CT image-based COVID-19 classification and analyzed the performance extensively in terms of accuracy, precision, recall, and AUC. Secondly, our study illustrates that misclassification can occur with a very minor perturbation magnitude, such as 0.009 and 0.003 for the FGSM attack in these models for X-ray and CT images, respectively, without any effect on the visual perceptibility of the perturbation. In addition, we demonstrated that successful FGSM attack can decrease the classification performance to 16.67% and 55.56% for X-ray images, as well as 36% and 40% in the case of CT images for VGG16 and Inception-v3, respectively, without any human-recognizable perturbation effects in the adversarial images. Finally, we analyzed that correct class probability of any test image which is supposed to be 1, can drop for both considered models and with increased perturbation; it can drop to 0.24 and 0.17 for the VGG16 model in cases of X-ray and CT images, respectively. Thus, despite the need for data sharing and automated diagnosis, practical deployment of such program requires more robustness.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: COVID-19 diagnosis using AI · Bacillus and Francisella bacterial research · Anomaly Detection Techniques and Applications
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Computer Science & Engineering, Rajshahi University of Engineering & Technology, Rajshahi 6204, Bangladesh
Department of Computer Science and Engineering, University of Rajshahi, Rajshahi 6205, Bangladesh
Department of Mathematics and Statistics, Imam Mohammad Ibn Saud Islamic University, Riyadh 13318, Saudi Arabia
WHO Collaborating Centre on eHealth, School of Public Health and Community Medicine UNSW Sydney, Sydney, NSW 2052, Australia
Healthy Ageing Theme, The Garvan Institute of Medical Research, Darlinghurst, NSW 2010, Australia
References (36)
Show source-provided references
1. Detection of SARS-CoV-2 in different types of clinical specimens
2. Zheng, C., Deng, X., Fu, Q., Zhou, Q., Feng, J., Ma, H., Liu, W., and Wang, X. (2020). Deep learning-based detection for COVID-19 from chest CT using weak label. MedRxiv.10.1101/2020.03.12.20027185 ↗
3. Sensitivity of chest CT for COVID-19: Comparison to RT-PCR10.1148/radiol.2020200432 ↗
4. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China10.1016/S0140-6736(20)30183-5 ↗
5. CT imaging of the 2019 novel coronavirus (2019-nCoV) pneumonia10.1148/radiol.2020200236 ↗
6. Emerging 2019 novel coronavirus (2019-nCoV) pneumonia10.1148/radiol.2020200274 ↗
7. Imaging profile of the COVID-19 infection: Radiologic findings and literature review
8. Correlation of chest CT and RT-PCR testing forcoronavirus disease 2019 (COVID-19) in China: A report of 1014 cases10.1148/radiol.2020200642 ↗
9. FDA (2021, April 15). FDA Permits Marketing of Artificial Intelligence-Based Device to Detect Certain Diabetes-Related Eye Problems. Silver Spring, DM, Department of Health and Human Services 2018, Available online: https://www.fda.gov/news-events/press-announcements/fda-permits-marketing-artificial-intelligence-based-device-detect-certain-diabetes-related-eye.
10. End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography10.1038/s41591-019-0447-x ↗
11. Overview of deep learning in medical imaging10.1007/s12194-017-0406-5 ↗
12. Gozes, O., Frid-Adar, M., Greenspan, H., Browning, P.D., Zhang, H., Ji, W., Bernheim, A., and Siegel, E. (2020). Rapid AI Development Cycle for the Coronavirus (covid-19) Pandemic: Initial Results for Automated Detection and Patient Monitoring Using Deep Learning CT Image Analysis. arXiv, Available online: https://arxiv.org/ftp/arxiv/papers/2003/2003.05037.pdf.
13. Li, L., Qin, L., Xu, Z., Yin, Y., Wang, X., Kong, B., Bai, J., Lu, Y., Fang, Z., and Song, Q. (2020). Artificial intelligence distinguishes COVID-19 from community acquired pneumonia on chest CT. Radiology.10.1148/radiol.2020200905 ↗
14. Wang, S., Kang, B., Ma, J., Zeng, X., Xiao, M., Guo, J., Cai, M., Yang, J., Li, Y., and Meng, X. (2020). A deep learning algorithm using CT images to screen for Corona Virus Disease (COVID-19). MedRxiv.10.1101/2020.02.14.20023028 ↗
15. Narin, A., Kaya, C., and Pamuk, Z. (2020). Automatic Detection of Coronavirus Disease (covid-19) Using X-ray Images and Deep Convolutional Neural Networks. arXiv, Available online: https://arxiv.org/ftp/arxiv/papers/2003/2003.10849.pdf.
16. Makris, A., Kontopoulos, I., and Tserpes, K. (2020, January 2–4). COVID-19 detection from chest X-ray images using Deep Learning and Convolutional Neural Networks. Proceedings of the 11th Hellenic Conference on Artificial Intelligence, Athens, Greece.10.1145/3411408.3411416 ↗
17. Karim, M., Döhmen, T., Rebholz-Schuhmann, D., Decker, S., Cochez, M., and Beyan, O. (2020). DeepCOVIDExplainer: Explainable Covid-19 Predictions Based on Chest X-ray Images. arXiv, Available online: https://arxiv.org/pdf/2004.04582.pdf.10.1109/BIBM49941.2020.9313304 ↗
18. Asnaoui, K.E., Chawki, Y., and Idri, A. (2020). Automated methods for detection and classification pneumonia based on X-ray images using deep learning. arXiv, Available online: https://arxiv.org/ftp/arxiv/papers/2003/2003.14363.pdf.
19. Goodfellow, I.J., Shlens, J., and Szegedy, C. (2014). Explaining and harnessing adversarial examples. arXiv, Available online: https://arxiv.org/pdf/1412.6572.pdf.
20. Simon, G. (1971). Principles of Chest X-ray Diagnosis, Butterworths.
21. Papernot, N., McDaniel, P., Wu, X., Jha, S., and Swami, A. (2016, January 23–25). Distillation as a defense to adversarial perturbations against deep neural networks. Proceedings of the 2016 IEEE Symposium on Security and Privacy (SP), San Jose, CA, USA.10.1109/SP.2016.41 ↗
22. Grosse, K., Papernot, N., Manoharan, P., Backes, M., and McDaniel, P. (2017, January 11–15). Adversarial Examples for Malware Detection. Proceedings of the 22nd European Symposium on Research in Computer Security, Oslo, Norway.10.1007/978-3-319-66399-9_4 ↗
23. Kurakin, A., Goodfellow, I., and Bengio, S. (2021, April 15). Adversarial Examples in the Physical World. Available online: https://arxiv.org/pdf/1607.02533.pdf.
24. Melis, M., Demontis, A., Biggio, B., Brown, G., Fumera, G., and Roli, F. (2017, January 22–29). Is deep learning safe for robot vision? Adversarial examples against the icub humanoid. Proceedings of the IEEE International Conference on Computer Vision Workshops, Venice, Italy.10.1109/ICCVW.2017.94 ↗
25. Cohen, J.P., Morrison, P., Dao, L., Roth, K., Duong, T.Q., and Ghassemi, M. (2020). Covid-19 Image Data Collection: Prospective Predictions Are the Future. arXiv, Available online: https://arxiv.org/pdf/2006.11988.pdf.10.59275/j.melba.2020-48g7 ↗
26. Yang, X., He, X., Zhao, J., Zhang, Y., Zhang, S., and Xie, P. (2020). COVID-CT-Dataset: A CT Scan Dataset about COVID-19. arXiv.
27. Pal, B., and Ahmed, B. (2016, January 20–22). A deep domain adaption approach for object recognition using Multiple Model Consistency analysis. Proceedings of the 2016 9th International Conference on Electrical and Computer Engineering (ICECE), Dhaka, Bangladesh.10.1109/ICECE.2016.7853982 ↗
28. O’Shea, K., and Nash, R. (2015). An Introduction to Convolutional Neural Networks. arXiv, Available online: https://arxiv.org/pdf/1511.08458.pdf.
29. Simonyan, K., and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. arXiv, Available online: https://arxiv.org/pdf/1409.1556.pdf.
30. Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., and Wojna, Z. (2016, January 27–30). Rethinking the inception architecture for computer vision. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, NV, USA.10.1109/CVPR.2016.308 ↗
31. Nguyen, A., Yosinski, J., and Clune, J. (2015, January 7–12). Deep neural networks are easily fooled: High confidence predictions for unrecognizable images. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Boston, MA, USA.10.1109/CVPR.2015.7298640 ↗
32. Tramer, F., Carlini, N., Brendel, W., and Madry, A. (2020). On Adaptive Attacks to Adversarial Example Defenses. arXiv, Available online: https://arxiv.org/pdf/2002.08347.pdf.
33. Adversarial robustness study of convolutional neural network for lumbar disk shape reconstruction from MR images
34. Adversarial examples: Attacks and defenses for deep learning10.1109/TNNLS.2018.2886017 ↗
35. Lu, J., Issaranon, T., and Forsyth, D. (2017, January 22–29). Safetynet: Detecting and rejecting adversarial examples robustly. Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy.10.1109/ICCV.2017.56 ↗
36. Ma, L., and Liang, L. (2021). Increasing-Margin Adversarial (IMA) Training to Improve Adversarial Robustness of Neural Networks. arXiv, Available online: https://arxiv.org/pdf/2005.09147.pdf.