Quantum-resistant blockchain-assisted federated secure cancer diagnosis framework using hierarchical window pattern cognition network
Abstract
Breast cancer is still a major cause of cancer mortality in women, and the development of efficient, reliable and explainable diagnostic systems capable of inter-hospital data sharing while maintaining privacy of data is essential. This study proposes a Hierarchical Window Pattern Cognition Network for breast cancer diagnosis using the Breast Histopathology Images (IDC Regular Patches) dataset. The workflow starts with data collection, then the Patient-Level Data Partitioning process by applying patient identifiers to avoid training-testing overlap. The framework uses CLAHE and data augmentation, image normalization and image resizing to enhance the image consistency. The network learns hierarchical histological and cellular morphological patterns by the modules: Global Histological Representation, Cellular Morphology Learning, Adaptive Tissue Fusion and Diagnostic Probability Estimation. A federated learning framework with five clients uses Federated Averaging (FedAvg) to aggregate models and share model updates securely. The security layer uses an AES-256-GCM encryption algorithm, a SHA3-512 hashing algorithm, and a mechanism simulating those of CRYSTALS-Kyber and CRYSTALS-Dilithium for cryptographic security. Record-integrity verification is supported by a local hash-chain audit log, and predictions of the model are explained with visual and feature-level explanations by using GradCAM + + and SHAP. The proposed centralized model achieved an accuracy of 94.23%, precision of 94.00%, recall of 94.47%, F1-score of 94.24%, and ROC-AUC of 0.9651. These findings highlight the promise of combining the histopathology classification, privacy-preserving federated learning, cryptographic protection, tamper-evident audit logging and explainable artificial intelligence within a unified diagnostic framework.
Related
- FFT-GA-BC: a fused federated transformer genetic algorithm framework with blockchain-secured, differentially private, homomorphically encrypted aggregation for mutation, phenotype, and binary-allele analysis.
- PoSFed: stake-aware validator selection and update filtering for blockchain-coordinated healthcare federated learning
- Federated learning with update level noise perturbation and blockchain audit logging for privacy-preserving healthcare ECG classification
- A federated attention-based stacked LSTM framework for interpretable malaria diagnosis under simulated non-IID federated conditions
- HeRZ: Hierarchical Recursive Zero-Knowledge Verification for Canonical and Auditable Asset-State Evolution in Blockchain-Cyber-Physical Systems
BibTeX
@article{asmeen2026quantumresistant,
title = {{Quantum-resistant blockchain-assisted federated secure cancer diagnosis framework using hierarchical window pattern cognition network}},
author = {Shaik Asmeen and Venkata Bhikshapathi Chenam},
journal = {Scientific Reports},
year = {2026},
month = nov,
doi = {10.1038/s41598-026-78779-6},
url = {https://www.nature.com/articles/s41598-026-78779-6},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Quantum-resistant blockchain-assisted federated secure cancer diagnosis framework using hierarchical window pattern cognition network
AU - Shaik Asmeen
AU - Venkata Bhikshapathi Chenam
PY - 2026
DA - 2026/11/07
JO - Scientific Reports
DO - 10.1038/s41598-026-78779-6
UR - https://www.nature.com/articles/s41598-026-78779-6
AB - Breast cancer is still a major cause of cancer mortality in women, and the development of efficient, reliable and explainable diagnostic systems capable of inter-hospital data sharing while maintaining privacy of data is essential. This study proposes a Hierarchical Window Pattern Cognition Network for breast cancer diagnosis using the Breast Histopathology Images (IDC Regular Patches) dataset. The workflow starts with data collection, then the Patient-Level Data Partitioning process by applying patient identifiers to avoid training-testing overlap. The framework uses CLAHE and data augmentation, image normalization and image resizing to enhance the image consistency. The network learns hierarchical histological and cellular morphological patterns by the modules: Global Histological Representation, Cellular Morphology Learning, Adaptive Tissue Fusion and Diagnostic Probability Estimation. A federated learning framework with five clients uses Federated Averaging (FedAvg) to aggregate models and share model updates securely. The security layer uses an AES-256-GCM encryption algorithm, a SHA3-512 hashing algorithm, and a mechanism simulating those of CRYSTALS-Kyber and CRYSTALS-Dilithium for cryptographic security. Record-integrity verification is supported by a local hash-chain audit log, and predictions of the model are explained with visual and feature-level explanations by using GradCAM + + and SHAP. The proposed centralized model achieved an accuracy of 94.23%, precision of 94.00%, recall of 94.47%, F1-score of 94.24%, and ROC-AUC of 0.9651. These findings highlight the promise of combining the histopathology classification, privacy-preserving federated learning, cryptographic protection, tamper-evident audit logging and explainable artificial intelligence within a unified diagnostic framework.
SN - 2045-2322
ER -