International Conference on Pervasive Computing and Social Networking (ICPCSN)

Toward Quantum-Enhanced Privacy in Medical Data Analytics: A Framework for QOTP-based Secure Computation

B. Gohin, M Ramabhadra Rao, K Lakshman Rao, Suresh Chanamala, R Sunitha, Y S Swarupa

Published May 6, 2026• 0 citations

Abstract

As quantum computing advances toward practical deployment, healthcare faces a dual challenge: enabling powerful computational analytics while guaranteeing rigorous patient data privacy. Existing classical approaches, particularly Fully Homomorphic Encryption (FHE), suffer from prohibitive performance costs and computational security assumptions vulnerable to quantum adversaries. This paper presents QPrivMed, a novel privacy-preserving framework for medical data computation to Noisy Intermediate-Scale Quantum (NISQ) devices. QPrivMed couples the Quantum One-Time Pad (QOTP) with selective homomorphic evaluation over Clifford circuits, achieving information-theoretic security for targeted clinical tasks without fault-tolerant hardware. Validated on 8-qubit IBM Quantum processors, the framework achieves 94.7% average circuit fidelity across binary medical classification scenarios, while outperforming classical FHE by a factor of 4.4× in execution time for equivalent Clifford workloads. Security analysis confirms zero mutual information leakage and tamper detection with probability ≥ 1 − 2−λ. This work bridges quantum cryptographic theory and practical healthcare data processing, establishing a scalable foundation for quantum-secured medical analytics while providing an honest assessment of current limitations and a realistic deployment timeline extending to the 2030.

DOI: 10.1109/icpcsn68523.2026.11544001

Related papers

Enhancing Healthcare Data Security Using Quantum Homomorphic Encryption

Sivaranjani Reddi, Jami Venkata Suman

This chapter presents a novel approach to securing sensitive healthcare data using quantum homomorphic encryption (QHE) integrated with the RSA encryption scheme. In an era where privacy-preserving data analysis is critical, particularly in healthcare, this method demonstrates how encrypted medical datasets, specifically the Diabetes dataset, can undergo computations such as homomorphic addition without compromising confidentiality. The chapter simulates encryption, quantum circuit-based homomorphic operations, and decryption using classical and quantum paradigms. It compares computational accuracy and efficiency between quantum and classical homomorphic models, establishing the feasibility of QHE for real-world medical applications. Performance is evaluated across multiple input sizes, demonstrating secure computation scalability. The results show that QHE offers promising capabilities for privacy-preserving analytics in healthcare and other sensitive domains, bridging quantum and classical cryptographic methods to support secure, efficient, and scalable data processing.

CRC Press eBooksNov 28, 20250 citations

Quantum Homomorphic Encryption: Towards Practical QOTP-Encrypted Computation on Gate-Based Hardware

Jon Hernández-Bueno, Oscar Lage, Marivi Higuero +1 more

As quantum computing moves toward remote and cloud-based access, protecting quantum data during outsourced execution becomes important. Quantum one-time pad (QOTP) encryption provides information-theoretic hiding of quantum state, but evaluating gates on Pauli-encrypted data is nontrivial: Clifford operations allow classical Pauli-frame updates, whereas non-Clifford operations require key-dependent corrections. We introduce QOTPH, a client-side key-tracked compilation framework for encrypted quantum computation. The client retains the QOTP keys and transforms each operation into a corrected gate or decomposition. The resulting circuit is executed non-adaptively by the quantum server, which neither receives the keys nor performs key updates. We develop a unified Pauli frame compilation layer spanning Clifford operations, one- and two-qubit Pauli-generated rotations, explicit key-dependent parametrized controlled rotations, and recursively composed gates, and establish a global correctness result for circuits composed of the formally verified primitive rules. We also characterize the evaluator's view under a hidden logical circuit model, in which the QOTP frame, the uncompiled logical circuit, and the logical-to-physical instruction mapping remain client-side; the security of this model against various algorithms is also analyzed. The implementation is evaluated in Qiskit using noiseless simulation and IBM Quantum hardware, including in-circuit and local-decryption workflows. This work bridges the gap between algebraic QOTP key tracking and practical encrypted quantum computation by providing a unified, hardware-validated compilation framework for current gate-based quantum processors.

quant-pharXiv:2604.19256Apr 22, 20260 citations

A Comparative Framework for Post-Quantum Cryptography and Homomorphic Encryption: Integration, Methodologies, and Applications

Dr Rajendirakumar S

Abstract The rapid evolution of quantum computing fundamentally threatens modern cryptographic infrastructures, necessitating a paradigm shift toward advanced encryption techniques. While post-quantum cryptography (PQC) ensures that data remains secure against quantum algorithmic decryption, it does not inherently facilitate the processing of that data in untrusted environments. Homomorphic encryption (HE) addresses this computational privacy requirement, yet traditional HE schemes rely on classical mathematical problems that are easily compromised by quantum systems. This paper presents a comprehensive comparative framework that integrates post-quantum cryptographic principles with homomorphic capabilities to enable quantum-resistant, privacy-preserving computation. By analyzing lattice-based algorithms, code-based cryptography, and hardware acceleration paradigms, this study maps the existing trade-offs between security, computational overhead, and practical deployment feasibility in the impending quantum era. Keywords: PQC, HE. 1.Introduction The rapid evolution of quantum computing technology presents a profound threat to modern cryptographic infrastructures. As quantum algorithms, notably Shor's algorithm, mature, they acquire the capability to solve nondeterministic polynomial time problems in polynomial time, thereby breaking traditional asymmetric encryption standards such as RSA and Elliptic Curve Cryptography (ECC) (Chen, 2024). Consequently, the development of post-quantum cryptography (PQC) has become a global imperative to secure sensitive communications and data against future quantum adversaries (Pranjal & Chaturvedi, 2024). Cryptographers are currently focusing on transitioning secure systems to these newly established PQC standards before large-scale quantum computers become commercially viable. While PQC addresses the secure transmission and storage of data, it does not inherently solve the problem of secure data processing in untrusted cloud environments. Homomorphic encryption (HE) provides a sophisticated mechanism that allows computations to be performed directly on encrypted data without requiring prior decryption (Jain & Cherukuri, 2023). The core problem addressed in this paper is the conceptual and practical integration of PQC with HE shown in Fig1.1 which is essential for ensuring quantum-safe, privacy-preserving computation in distributed architectures. Achieving this synthesis is highly complex due to the massive computational overhead and noise accumulation inherent to both cryptographic domains. However, existing approaches to privacy-preserving computation remain broadly insufficient for the post-quantum era. First, conventional homomorphic encryption schemes rely heavily on mathematical hardness assumptions, such as integer factorization or discrete logarithms, which are demonstrably vulnerable to quantum attacks (Chen, 2024). Second, while some high-performance fully homomorphic implementations exist, they suffer from prohibitive computational overhead and severe ciphertext expansion, making widespread deployment impractical without specialized mitigation strategies (Tseng et al., 2025). Third, hardware-accelerated privacy solutions often rely on trusted execution environments, like Intel SGX, which inherently introduce side-channel vulnerabilities and force reliance on centralized hardware trust rather than pure mathematical security (Sadat et al., 2017). To overcome these critical limitations, this paper proposes a structured comparative methodology for evaluating quantum-resistant homomorphic encryption systems. Specifically, our paper makes the following main contributions: We provide a comprehensive comparative framework that categorizes and evaluates lattice-based and code-based homomorphic encryption schemes against classical performance constraints. We propose a structured, hypothetical evaluation pipeline designed to benchmark the computational overhead and latency of post-quantum homomorphic algorithms in federated networking environments. 2.Privacy-preserving computation The literature surrounding privacy-preserving computation can be broadly classified into three primary categories, the first being classical and fully homomorphic encryption (FHE) systems. The core idea behind these schemes is the preservation of algebraic structures during encryption, allowing operations on ciphertexts to reflect directly on the underlying plaintexts without exposing the raw data (Jain & Cherukuri, 2023). A major strength of these classical FHE systems, such as those utilized in encrypted relational databases like FHE-SQL, is their ability to provide end-to-end cryptographic protection for queries without relying on trusted hardware modules (Tseng et al., 2025). However, a fundamental weakness is that early FHE models, and their underlying privacy homomorphisms, remain fundamentally susceptible to quantum cryptanalysis due to their reliance on classical mathematical problems (Sen, 2013). Compared to these classical paradigms, our work specifically focuses on migrating these operational homomorphic properties into genuinely quantum-safe mathematical frameworks. The second category of related work focuses on hardware acceleration and hybrid computational environments for homomorphic encryption. The core idea here involves leveraging specialized hardware, such as ASIC AI chips or coarse-grained compute architectures like Google's TPUs, to mitigate the enormous computational overhead associated with HE (Tong et al., 2025). The strength of this approach lies in its ability to bridge the energy-efficiency gap, often utilizing intelligent compiler frameworks to convert high-precision modular arithmetic into dense, low-precision matrix multiplications. Conversely, the weakness of hardware-centric methods is that architectural mismatches frequently lead to severe resource under-utilization, and hybrid solutions utilizing components like Intel SGX introduce hardware trust assumptions that can compromise pure cryptographic integrity (Sadat et al., 2017). Unlike these approaches, our framework prioritizes algorithmic resilience and quantum resistance over purely hardware-driven and enclave-based optimizations. The third category encompasses emerging post-quantum and quantum homomorphic paradigms, which seek to unify quantum resistance with computational privacy. Researchers have proposed lattice-based HE schemes to replace vulnerable legacy systems like Paillier and ElGamal, successfully enabling secure applications such as federated learning (Chen, 2024). Similarly, code-based cryptography utilizing McEliece and non-negative matrix factorization has shown promise in providing faster encryption and decryption times while remaining mathematically secure against quantum computers (Chen, 2024). Furthermore, advanced quantum homomorphic encryption protocols have explored information-theoretic security combined with homomorphic quantum error-correction to ensure future-proof data protection (Ouyang & Rohde, 2022). While these systems represent the state-of-the-art, their primary weakness remains the massive memory and communication overhead introduced during distributed training protocols, which negatively impacts classification performance. Our approach builds upon this category by proposing a standardized methodology to systematically compare these disparate post-quantum homomorphic techniques. 3.Multi-stage Method To systematically evaluate and deploy post-quantum homomorphic encryption, we propose a structured, multi-stage methodological framework. The first step involves cryptographic primitive selection, where system architects must evaluate and choose between lattice-based constructs, such as ML-KEM adaptations, and code-based systems like McEliece, depending on their domain constraints (Chen, 2024). The second step requires the implementation of an algorithmic parameter optimization module, which dynamically adjusts polynomial degrees and error distributions to balance security levels with the operational multiplicative depth required by the application. Finally, the third step comprises the integration of a specialized abstraction layer designed to map complex high-precision cryptographic operations into hardware-friendly matrix operations, thereby reducing the inherent bottleneck of ciphertext evaluation. The key design choices within this framework are driven by the absolute necessity to balance quantum resistance with practical algorithmic utility. We prioritize lattice-based cryptography for our primary evaluation pathway because of its established resilience against quantum threats and its recent standardization by the National Institute of Standards and Technology (NIST) (Chen, 2024). Additionally, we explicitly design our framework to avoid reliance on trusted hardware environments, opting instead for pure cryptographic guarantees to prevent the side-channel data leaks that often plague hardware-enclave solutions. By focusing on algorithmic alignment and basis-aligned mathematical transformations, the framework ensures that homomorphic workloads can be optimized effectively without compromising the underlying post-quantum security proofs. To validate the efficacy of the proposed framework, we outline a comprehensive evaluation plan utilizing a hypothetical, high-dimensional healthcare benchmarking dataset. The experimental setup will simulate a secure genome-wide association study (GWAS) executing within a fully decentralized federated learning environment (Sadat et al., 2017). The primary metrics for this hypothetical evaluation will include key generation latency, ciphertext expansion ratio, and the amortized computational time required for homomorphic multiplication operations across distributed nodes. By systematically varying the cryptographic noise parameters and observing the resulting degradation in analytical performance, we aim to empirically map the trad

Dec 23, 20250 citations

Advancing Healthcare Using Quantum Computing and Federated Learning: Challenges and Future Perspectives

Shilpa Ankalaki, Geetabai S Hukkeri, Shaleen Bhatnagar

The increasing availability of electronic health records, medical imaging databases, and sensor-generated clinical datasets has boosted the need for privacy-preserving yet scalable analytical frameworks. In this regard, quantum computing and federated learning are becoming two focal points that may significantly shape future healthcare data processing. Quantum computing provides powerful tools for working with complex, high-dimensional optimization and learning problems, while federated learning facilitates collaborative model training without requiring the centralized exchange of sensitive information about patients. This review explores the integration of these two technologies and Quantum Split Federated Learning (QSFL) as a promising approach for secure and scalable co-operative health care analytics. The treatment then turns to discuss the quantum computation mathematical and algorithmic basics, and subsequently, quantum machine learning methods based on extremes of supervised, unsupervised, and semi-supervised modes. Special emphasis is placed on how QSFL can enable privacy-aware inter-institutional learning and enhance clinical decision support systems. It also surveys the applications for creating high-quality synthetic medical data sets for training, validation and data-sharing using quantum generative adversarial networks (qGANs). In addition to the technological opportunities this paper critically evaluates the practical challenges to real-world implementation, such as hardware limitations, communication costs, regulatory matters, and ethical considerations of a distributed medical learning system-based infrastructure and mechanism. In the end, the survey identifies important open research problems and provides suggestions for future research directions towards realizing quantum-enabled, privacy-preserving healthcare intelligence systems that are anticipated to usher in a new era of medical analytics.

Archives of Computational Methods in EngineeringMar 31, 20262 citations

Hybrid Quantum-Classical Computing for Privacy-Preserving Machine Learning on Edge Devices

B. Dinesh, M. Thangamani, B. Paulchamy +3 more

The edge computing environment faces a challenge in maintaining data privacy while performing machine learning tasks. The current solutions for privacy preservation are inefficient or do not provide security guarantees, causing a critical gap in the edge computing environment, which is resource-constrained. The aim of this research work is to design a hybrid quantum-classical computing system that addresses the challenges of privacy preservation and efficiency in machine learning tasks. We design a novel architecture combining variational quantum circuits for secure data encoding with classical neural networks for scalable inference, implemented through federated learning protocols enhanced with quantum differential privacy mechanisms. Experimental evaluation demonstrates a 35% reduction in communication overhead compared to traditional federated learning approaches while maintaining 99.2% classification accuracy across diverse datasets. The quantum-enhanced privacy mechanisms achieve provable security guarantees with information-theoretic bounds, outperforming conventional cryptographic methods by 28% in computational overhead reduction. This hybrid approach establishes a new paradigm for secure and efficient machine learning deployment on edge devices, enabling privacy-preserving artificial intelligence in healthcare, finance, and IoT applications.

IEEE International Conference on Emerging Computing and Intelligent Technologies (ICoECIT)Jan 28, 20260 citations

MQFL-FHE: Multimodal Quantum Federated Learning Framework with Fully Homomorphic Encryption

Siddhant Dutta, Nouhaila Innan, Sadok Ben Yahia +2 more

The integration of fully homomorphic encryption (FHE) in federated learning (FL) has led to significant advances in data privacy. However, during the aggregation phase, it often results in performance degradation of the aggregated model, hindering the development of robust representational generalization. In this work, we propose a novel multimodal quantum federated learning framework that utilizes quantum computing to counteract the performance drop resulting from FHE. For the first time in FL, our framework combines a multimodal quantum mixture of experts (MQMoE) model with FHE, incorporating multimodal datasets for enriched representation and task-specific learning. Our MQMoE framework enhances performance on multimodal datasets and combined genomics and brain MRI scans, especially for underrepresented categories. Our results also demonstrate that the quantum-enhanced approach mitigates the performance degradation associated with FHE and improves classification accuracy across diverse datasets, validating the potential of quantum interventions in enhancing privacy in FL.

arXiv:2412.01858International Joint Conference on Neural Networks (IJCNN)Jun 30, 20253 citations