Citation :

B. Murugeshwari, Kayalvizhi S, Daya Florance D, Saranya S, "A Cryptographic Proof Framework for Tamper-Evident Pneumonia Diagnosis on Blockchain," International Journal of Electronics and Communication Engineering, vol. 13, no. 7, pp. 140-154, 2026. Crossref, https://doi.org/10.14445/23488549/IJECE-V13I7P110

Abstract

To deploy deep learning-based diagnostic systems in a clinical setting, they need to have not only a high degree of predictive accuracy, but also an unbroken cryptographic chain to prove that the model parameters have not been changed from one inference to the next. This incompatibility arises because softmax, GELU activation, and layer normalization involve transcendental operations to produce the final output. Transcendental operations cannot be represented over the prime finite fields that are necessary when utilizing Rank-1 Constraint Systems (R1CS). The work here provides a mathematically sound approach to resolving the incompatibility by providing three formally defined polynomial approximations: a degree-3 Taylor series softmax approximation (ε ≤ 2.47E-03 per weight); a degree-5 composite polynomial GELU (ε ≤ 1.80E-03); and a squared witness hint reformulation of layer normalization that results in exact constraint satisfaction. The accumulating polynomial approximation errors across 12 transformer encoder blocks have a total approximation error defined as (ε total ≤ 0.0512) and results in a loss of less than one-fifth of a percent in accuracy (94.1% vs. 94.3%) on the RSNA Pneumonia Detection benchmark. The entire ViT-B/16 inference model was compiled into an R1CS form of approximately 2.3×10⁸ R1CS constraints and instantiated as a Groth 16 zk-SNARK. The integrity of the model is confirmed by hashing the parameters using SHA-256 onto an immutable Polygon zkEVM smart contract, allowing for on-chain verification of inference without revealing any of the proprietary model weights. Additionally, this system produces an area under the receiver operating characteristic curve (AUC-ROC) of 0.961, a mean latency for proof generation of 2.84 seconds, an end-to-end verification time of 5.07 seconds, and an average cost for on-chain verification of 0.012ETH, all within the operational constraints of typical radiology workflows.

Keywords

Cryptographic chain, Polynomial approximation, Pneumonia diagnosis.

References

[1] Aitizaz Ali et al., “Blockchain-Powered Healthcare Systems: Enhancing Scalability and Security with Hybrid Deep Learning,” Sensors, vol. 23, no. 18, pp. 1-33, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[2] Duggineni Srinivasa Rao et al., “A Hybrid Deep Learning-Based Decision-Making Framework for Scalability and Security in Blockchain-Powered Healthcare Systems,” Frontiers in Health Informatics, vol. 13, no. 6, pp. 855-864, 2024.
[Google Scholar] [Publisher Link]

[3] Bowei Zhang et al., “A Blockchain and Zero-Knowledge Proof Based Data Security Transaction Method in Distributed Computing,” Electronics, vol. 13, no. 21, pp. 1-20, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[4] Shahinul Hoque et al., “Deep Learning Model Integrity Checking Mechanism Using Model Watermarking,” arXiv preprint, pp. 1-5 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[5] Xin Mu et al., “Model Provenance via Model DNA,” arXiv preprint, pp. 1-10, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[6] Malgorzata Schwab, and Ashis Kumer Biswas “Invertible Neural Networks for Trustworthy AI,” 2023 IEEE 35^th International Conference on Tools with Artificial Intelligence (ICTAI), Atlanta, GA, USA, pp. 480-485, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[7] Mikhail Pautov et al., “Probabilistically Robust Watermarking of Neural Networks,” Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence (IJCAI-24), Jeju, Korea, pp. 4778-4787, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[8] Tzu-Yun Chien, and Chih-Ya Shen, “Customized and Robust Deep Neural Network Watermarking,” Proceedings of the 17^th ACM International Conference on Web Search and Data Mining, Merida Mexico, pp. 134-142, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[9] Sarthak Kumar Maharana, Yuxuan Li, and Yunhui Guo “Not Just Change the Labels, Learn the Features: Watermarking Deep Neural Networks with Multi-view Data,” 18^th European Conference Computer Vision – ECCV 2024, Milan, Italy, pp. 273-289, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[10] Saoussen Ben Jabra, and Mohamed Ben Farah, “Deep Learning-Based Watermarking Techniques: Challenges and Trends,” Circuits, Systems, and Signal Processing, vol. 43, pp. 4339-4368, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[11] Chaoxiang He et al., “Towards Stricter Black-box Integrity Verification of Deep Neural Network Models,” Proceedings of the 32^nd ACM International Conference on Multimedia, Melbourne VIC Australia, pp. 9875-9884, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[12] Shayne Longpre et al., “Position: Data Authenticity, Consent, & Provenance for AI are all Broken: What will it Take to Fix them?,” Proceedings of the 41^st International Conference on Machine Learning Research Workshop, vol. 235, pp. 1-15, 2024.
[Google Scholar] [Publisher Link]

[13] Xiaofeng Liu et al., “Exploring Backdoor Attacks in Off-the-Shelf Unsupervised Domain Adaptation for Securing Cardiac MRI-Based Diagnosis,” 2024 IEEE International Symposium on Biomedical Imaging (ISBI), Athens, Greece, pp. 1-5, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[14] Weimin Lyu et al., “BadCLM: Backdoor Attacks in Clinical Language Models,” AMIA Annual Symposium Proceedings, pp. 1-12, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[15] Marius Schlegel, and Kai-Uwe Sattler, “Capturing End-to-End Provenance for Machine Learning Pipelines,” Information Systems, vol. 132, pp. 1-15, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[16] Carlos Melo et al., “Performance Modeling and Evaluation of Hyperledger Fabric: An Analysis Based on Transaction Flow and Endorsement Policies,” 2024 IEEE Symposium on Computers and Communications (ISCC), Paris, France, pp. 1-6, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[17] Sam Goundar, and Iqbal Gondal, “AI-Blockchain Integration for Real-Time Cybersecurity: System Design and Evaluation,” Sensors, vol. 5, no. 3, pp. 1-22, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[18] Shrawan Kumar Sharma, and Firoj Parwej, “Design and Implementation of a Blockchain-Based Secure Data-Sharing Framework for Healthcare,” Blockchains, vol. 3, no. 3, 1-25, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[19] Jiyu Yang et al., “EdgeThemis: Ensuring Model Integrity for Edge Intelligence,” Proceedings of the ACM on Web Conference 2025, Sydney NSW Australia, pp. 3136-3146, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[20] Shridhar Singh, and Luke Vorster, “LLM Supply Chain Provenance: A Blockchain-based Approach,” International Conference Proceedings on AI Research, Lisbon, Portugal, vol. 5, no 1, pp. 381-389, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[21] Tsung-Ting Kuo, Hyeon-Eui Kim, and Lucila Ohno-Machado, “Blockchain Distributed Ledger Technologies for Biomedical and Health Care Applications,” Journal of the American Medical Informatics Association, vol. 24, no. 6, pp. 1211-1220, 2017.
[CrossRef] [Google Scholar] [Publisher Link]

[22] Asaph Azaria et al., “MedRec: Using Blockchain for Medical Data Access and Permission Management,” 2016 2^nd International Conference on Open and Big Data, Vienna, Austria, pp. 25-30, 2016.
[CrossRef] [Google Scholar] [Publisher Link]

[23] Dinh C. Nguyen et al., “Blockchain for Secure EHRs Sharing of Mobile Cloud Based E-Health Systems,” IEEE Access, vol. 7, pp. 66792-66806, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[24] Jens Groth, “On the Size of Pairing-Based Non-interactive Arguments,” Advances in Cryptology – EUROCRYPT 2016: 35^th Annual International Conference on the Theory and Applications of Cryptographic Techniques, Vienna, Austria, pp. 305-326, 2016.
[CrossRef] [Google Scholar] [Publisher Link]

[25] Jiasi Weng et al., “pvCNN: Privacy-Preserving and Verifiable Convolutional Neural Network Testing,” IEEE Transactions on Information Forensics and Security, vol. 18, pp. 2218-2233, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[26] Daniel Kang et al., “Scaling up Trustless DNN Inference with Zero-Knowledge Proofs,” arXiv preprint, pp. 1-12, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[27] Theo Ryffel et al., “A Generic Framework for Privacy Preserving Deep Learning,” arXiv preprint, pp. 1-5, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[28] Georgios A. Kaissis et al., “Secure, Privacy-Preserving and Federated Machine Learning in Medical Imaging,” Nature Machine Intelligence, vol. 2, pp. 305-311, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[29] Gan Sun et al., “Data Poisoning Attacks on Federated Machine Learning,” IEEE Internet of Things Journal, vol. 9, no. 13, pp. 11365-11375, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[30] Ruchika Bhuria et al., “Automated Pneumonia Detection from Chest X-Rays Using Convolutional Neural Networks: Advancements, Optimization, and Performance Analysis,” 2024 2^nd International Conference on Sustainable Computing and Smart Systems (ICSCSS), Coimbatore, India, pp. 1360-1364, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[31] Alexey Dosovitskiy et al., “An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale,” International Conference on Learning Representations (ICLR), pp. 1-21, 2021.
[Google Scholar] [Publisher Link]

[32] Jieneng Chen et al., “TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation,” arXiv preprint, pp. 1-13, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[33] Ze Liu et al., “Swin Transformer: Hierarchical Vision Transformer using Shifted Windows,” 2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada, pp. 9992-10002, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[34] Eli Ben-Sasson et al., “SNARKs for C: Verifying Program Executions Succinctly and in Zero Knowledge,” Advances in Cryptology – CRYPTO 2013: 33^rd Annual Cryptology Conference, Santa Barbara, CA, USA, pp. 90-108, 2013.
[CrossRef] [Google Scholar] [Publisher Link]

[35] Kaiming He et al., “Deep Residual Learning for Image Recognition,” 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770-778, Las Vegas, NV, USA, 2016.
[CrossRef] [Google Scholar] [Publisher Link]

[36] Gao Huang et al., “Densely Connected Convolutional Networks,” 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, pp. 2261-2269, 2017.
[CrossRef] [Google Scholar] [Publisher Link]

[37] Mingxing Tan, and Quoc Le, “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks,” Proceedings of the 36^th International Conference on Machine Learning, vol. 97, pp. 6105-6114, 2019.
[Google Scholar] [Publisher Link]