DPD Helix quantum inspired multi biometric medical image watermarking framework using Quantum Particle Swarm Optimization

Abstract

Abstract The growing reliance on digital medical imaging and Electronic Health Records (EHR) has heightened the need for robust mechanisms to ensure patient data authenticity, integrity, and confidentiality. This paper presents DPD-Helix, a quantum-inspired multi-biometric medical image watermarking framework that synergistically integrates Dual-Tree Complex Wavelet Transform (DTCWT), Pseudo-Zernike Moments (PZM), DNA encoding, and Quantum Particle Swarm Optimization (QPSO). DTCWT provides shift-invariant multi-resolution decomposition across 18 directional sub-bands, while PZM enables rotation-invariant, content-adaptive identification of perceptually optimal embedding locations using an empirically validated threshold (τ = 0.85). Dual-biometric watermarks comprising fingerprint (FVC2004) and iris (MMU) templates are fused through key-dependent DNA XOR operations, ensuring cryptographic security grounded in biological computing principles. QPSO, modelling particles as quantum entities within delta potential wells, exploits superposition and tunnelling effects to achieve globally optimal embedding strength parameters, converging 57.7% faster than classical PSO with a 98.9% global optima discovery rate. Comprehensive experiments on three Kaggle medical imaging datasets which include Brain MRI (3264 images), Chest X-ray (5863 images), and Retinal OCT (84,495 images) were conducted using MATLAB R2024b. Results demonstrate that DPD-Helix achieves a mean PSNR of 52.83 dB, SSIM of 0.9989, and NC exceeding 0.9920 across 20 attack scenarios, including JPEG compression, Gaussian noise, rotation, scaling, and cropping. Comparative analysis confirms that DPD-Helix outperforms six state-of-the-art methods, with formal security analysis establishing a key space of 1.47 × 10 26 and validated DICOM/PACS compatibility for clinical deployment in telemedicine and medical image archiving systems.