Counterfeit Drug Detection Using Deep Learning and Synthetic Data Generation: A Hybrid Approach
DOI:
https://doi.org/10.67378/1s0a6r05Keywords:
Counterfeit drug detection; deep learning; synthetic data generation; diffusion models; computer vision; pharmaceutical authentication; transfer learning.Abstract
Counterfeit pharmaceuticals pose a critical threat to global public health, particularly in low- and middle-income countries where regulatory oversight and access to advanced authentication tools are limited. This study presents a comprehensive deep learning framework for automated counterfeit drug detection that addresses the persistent challenge of limited authentic counterfeit training data through synthetic data generation. A hybrid dataset was constructed by combining real-world pharmaceutical images from public repositories with synthetically generated counterfeit images produced using Stable Diffusion XL. Five convolutional neural network architectures—ResNet50, InceptionV3, VGG16, EfficientNetB0, and MobileNetV2—were evaluated under real-only, synthetic-only, and hybrid training regimes. The hybrid approach achieved the best performance, with ResNet50 attaining 97.0% accuracy and an AUC-ROC of 0.993, significantly outperforming models trained on real data alone. Integration of optical character recognition for batch number and expiration date verification further improved system accuracy to 98.2%. A mobile application prototype was developed to demonstrate deployment feasibility, achieving sub-200 ms inference time after model optimization. The results confirm that carefully designed synthetic data generation, when combined with real-world data, substantially improves robustness, generalization, and practical usability of AI-based pharmaceutical authentication systems.
References
[1] World Health Organization, “Substandard and falsified medical products,” WHO Fact
Sheet, 2023. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/substandard- and-falsified-medical-products
[2] World Health Organization, “WHO Global Surveillance and Monitoring System for sub-
standard and falsified medical products,” Geneva, Switzerland, WHO Technical Report, 2022.
[3] International Trademark Association, “The economic impact of counterfeiting and piracy:
Report prepared for BASCAP and INTA,” Frontier Economics, London, UK, 2021.
[4] Dégardin, K., Roggo,Y., and Margot, P.(2014). Understanding and fighting the medicine DOI: https://doi.org/10.1016/j.jpba.2013.01.009
counterfeit market. Journal of Pharmaceutical and Biomedical Analysis, vol. 87, pp. 167-175.
[5] Bansal, A., Garg, L. and Sharma, R. (2020). Artificial intelligence in pharmaceutical
industry: A review. Journal of Pharmaceutical Sciences and Research, vol. 12, no. 3, pp. 412-421, 2020.
[6] LeCun, Y., Bengio, Y. and Hinton, G. (2015). Deep learning. Nature, vol. 521, no. 7553, pp. DOI: https://doi.org/10.1038/nature14539
436-444.
[7] Mackey, T. and Nayyar, G. (2017). A review of existing and emerging digital technologies
to combat the global trade in fake medicines. Expert Opinion on Drug Safety, vol. 16, no. 5, pp. 587-602.
[8] Newton, P., Green, M. and Fernández, F. (2010). Impact of poor-quality medicines in the DOI: https://doi.org/10.1016/j.tips.2009.11.005
de- veloping’ world,” Trends in Pharmacological Sciences, vol. 31, no. 3, pp. 99-101.
[9] Shorten, A. and Khoshgoftaar, T. (2019). A survey on image data augmentation for deep DOI: https://doi.org/10.1186/s40537-019-0197-0
learning. Journal of Big Data, vol. 6, no. 1, pp. 1-48.
[10] Alsallal, M., Thompson, K. and Zaid, H. (2020). Counterfeit drug detection using X-ray
fluores- cence and machine learning. Journal of Pharmaceutical Sciences, vol. 109, no. 8, pp. 2553-2561.
[11] Mondal, S., Wijewardena, K. and Karuppuswami, S. (2021). MediVerify: A computer
vision and blockchain platform for drug authentication. in Proc. IEEE International Conference on Blockchain and Cryptocurrency (ICBC), Sydney, Australia, pp. 1-5
[12] Kumar, R. Patel, S. and Singh, V. (2021). Multi-modal pharmaceutical authentication using
OCR and CNN. Expert Systems with Applications, vol. 185, Article ID 115634.
[13] Goodfellow, I. (2014). Generative adversarial nets. in Advances in Neural Information Processing
Systems (NeurIPS), Montreal, Canada, pp. 2672-2680.
[14] Zhang, L. Zhu, W. and Wang, J(2020).Medical image synthesis using generative
adversarial networks: A review. IEEE Access, vol. 8, pp. 185734-185753.
[15] Ho, J., Jain, A.and Abbeel, P. (2020). Denoising diffusion probabilistic models. in Advances in
Neural Information Processing Systems (NeurIPS), Virtual Conference, pp. 6840-6851.
[16] Rombach, R. Blattmann, A., Lorenz, D., Esser, P. and Ommer, B. (2022). High-resolution
image synthesis with latent diffusion models. in Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, LA, 2022, pp. 10684-0695.
[17] Patel, S. and Kumar, R.(2022.“Smartphone-captured pharmaceutical product images
dataset. Mendeley Data, v2. [Online]. Available: https://data.mendeley.com/datasets
[18] Fernando, H. (2023). Authentic and counterfeit drug images dataset. Kaggle. [Online]. Available:
https://www.kaggle.com/datasets
[19] Buslaev, A., Iglovikov, V., Khvedchenya, E., Parinov, A., Druzhinin, M. and Kalinin, A. (2020).
Albumentations: Fast and flexible image augmentations,” Information, vol. 11, no. 2, p. 125. DOI: https://doi.org/10.3390/info11020125
[20] Smith, R. (2007). An overview of the Tesseract OCR engine. in Proc. International Conference on DOI: https://doi.org/10.1109/ICDAR.2007.4376991
Document Analysis and Recognition (ICDAR), Curitiba, Brazil, pp. 629-633.
[21] Agarwal, A., Singh, P. and Gupta, M., (2022). Mobile application for counterfeit drug detection
using convolutional neural networks. Mobile Information Systems, vol. 2022, Article ID 5847293.
[22] Perez, L. and Wang, J. (2017). The effectiveness of data augmentation in image classification
using deep learning. arXiv preprint arXiv:1712.04621.
[23] Chawla, N., Bowyer, K., Hall, L. and Kegelmeyer, W. (2002). SMOTE: Synthetic minority
over-sampling technique. Journal of Artificial Intelligence Research, vol. 16, pp. 321-357.
[24] Krizhevsky, A., Sutskever, I., and Hinton, G. (2012). ImageNet classification with deep
convolu tional neural networks. in Advances in Neural Information Processing Systems (NeurIPS), Lake Tahoe, NV, pp. 1097-1105.
[25] Newton, P.,Green, M., Fernández, F., Day, N., and White, N. (2006). Counterfeit anti-infective drugs. The Lancet Infectious Diseases, vol. 6, no. 9, pp. 602-613. DOI: https://doi.org/10.1016/S1473-3099(06)70581-3
Downloads
Published
License
Copyright (c) 2026 Ivan Arabome, Evan O. Arabome (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright remains with the author(s). This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits use, sharing, adaptation, distribution, and reproduction in any medium or format, provided appropriate credit is given to the original author(s) and source.
