Content-Aware DNA-Coded Cryptosystem for Skin Lesion Image Confidentiality in IoMT

Abstract

Teledermatology screening pipelines routinely transmit patient dermoscopic skin lesion images across Internet of Medical Things (IoMT) networks, where diffusion strength is conventionally applied uniformly across an image without regard to local visual complexity, an inefficiency increasingly addressed in general IoT image security through lightweight CNN-based content classifiers but largely unexplored for dermoscopic imaging. This work proposes a content-aware alternative in which a lightweight Convolutional Neural Network (CNmost existing chaos-DNA encryption schemes apply a fixed diffusion strength...N) first analyses each image to extract entropy, texture, and edge-density descriptors; an adaptive decision module then converts this feature vector into an image-specific chaos-strength parameter and DNA mutation block level before any encryption occurs. Independently, an 8-qubit quantum circuit executed on Qiskit generates a true quantum random bitstream via Hadamard superposition and computational-basis measurement; this stream is fused with the image\'s own SHA-384 digest to derive a hybrid initial condition for a 3D Logistic-Sine hyperchaotic map, whose three output sequences are respectively assigned to spatial scrambling, DNA-level mutation, and XOR diffusion. Each RGB channel is processed independently through Haar-wavelet scrambling, adaptive Bio-DNA encoding and mutation, and quantum-keyed diffusion. Evaluated on 600 dermoscopic skin lesion images from the ISIC archive, the framework attains 7.9887-bit average entropy, 99.6128% NPCR, 33.4377% UACI, near-zero adjacent-pixel correlation (0.0161), and fully lossless decryption (MSE = 0) in an average of 0.64 seconds per image confirming that content-aware, CNN-guided diffusion strength is compatible with strong statistical security and real-time IoMT deployment.

Introduction

The text presents a content-adaptive encryption framework for securing dermoscopic medical images in teledermatology and IoMT healthcare systems. The motivation is that skin-lesion images contain sensitive patient information and are often transmitted through public or semi-trusted networks, making confidentiality, integrity, and authenticity essential.

Traditional encryption methods such as AES and DES can be computationally demanding for large medical images because of their high pixel correlation and data volume. To address this, the proposed system combines chaotic encryption, Bio-DNA encoding, quantum random number generation, and lightweight CNN-based content analysis.

Main Objective

The primary goal is to develop an encryption method that adapts its encryption strength to the complexity of each medical image. Instead of applying the same encryption parameters to every image, the system analyzes image characteristics and dynamically adjusts the encryption process.

Proposed Framework

The framework consists of several major stages:

• Image preprocessing: RGB dermoscopic images are resized to 128 × 128 pixels, normalized, and separated into Red, Green, and Blue channels.
• CNN-based feature analysis: A lightweight CNN analyzes image content and extracts descriptors related to entropy, texture, and edge density.
• Adaptive decision module: These features are combined into a complexity score that determines:
  • the strength of the chaotic system, and
  • the DNA mutation block size (8, 16, or 32 pixels).
• Quantum random number generation: An 8-qubit Qiskit circuit uses Hadamard gates and measurement to generate a random bitstream.
• SHA-384 hybrid seeding: The quantum random data is combined with the SHA-384 hash of the original image, making the cryptographic seed dependent on both quantum randomness and image content.
• 3D Logistic-Sine hyperchaotic system: The hybrid seed initializes the chaotic system, which generates sequences used for encryption.
• DWT-based scrambling: Haar Discrete Wavelet Transform is used to scramble image information in the transform domain.
• Bio-DNA encoding and mutation: DNA-inspired encoding and adaptive mutation increase confusion and diffusion.
• XOR diffusion: The generated chaotic sequences are used for further pixel-level diffusion.
• RGB recombination: Independently encrypted RGB channels are combined to form the final encrypted image.

Key Innovation

The main research gap identified is that many existing chaos-DNA and quantum-assisted image encryption systems use fixed encryption parameters regardless of image content. The proposed approach introduces a CNN-guided adaptive mechanism, allowing visually complex and information-rich images to receive stronger, finer-grained encryption while simpler images can use less computationally intensive settings.

Related Research

Previous research has explored:

• Chaos and hyperchaos for image scrambling and diffusion.
• DNA computing for complex image encoding and mutation.
• Quantum-inspired methods for stronger key generation.
• DWT and transform-domain encryption.
• AI-based segmentation and content-aware medical image protection.
• Hybrid quantum-classical encryption for IoMT systems.

However, the text argues that these approaches generally do not combine CNN-driven adaptive encryption strength, quantum random initialization, hyperchaotic encryption, and Bio-DNA processing in a single framework.

Applications

The proposed system is intended for:

• Teledermatology and remote skin-lesion diagnosis
• Cloud-based PACS/EHR systems
• Resource-constrained IoMT devices and gateways
• Secure medical research data sharing
• Real-time transmission of sensitive medical images

Conclusion

This paper presented an AI-Assisted, adaptive quantum-hyperchaotic Bio-DNA framework for securing dermoscopic skin lesion images in IoMT environments. A lightweight CNN extracts entropy, texture, and edge-density descriptors from each image; an adaptive decision module converts these into an image-specific chaos-strength parameter and DNA mutation block level; and a Qiskit-executed quantum circuit supplies a true random bitstream that is fused with the image\'s SHA-384 digest to seed a 3D Logistic-Sine hyperchaotic map driving wavelet scrambling, adaptive DNA mutation, and diffusion across each RGB channel independently. Evaluated on 600 dermoscopic skin lesion images, the framework achieved 7.9887-bit entropy, 99.6128% NPCR, 33.4377% UACI, near-zero pixel correlation, and fully lossless decryption in an average of 0.64 seconds confirming both strong statistical security and real-time practicality. Future work will target formal validation of the hyperchaotic map\'s dynamical behaviour (e.g., via Lyapunov exponent analysis), execution of the quantum random number generation stage on physical NISQ-era hardware rather than a simulator, extension to full-resolution multi-modality medical imaging (MRI, CT), and deployment validation on constrained IoMT edge platforms.

References

[1] M. S. Khan, A. Al-Dubai, J. Ahmad, N. Pitropakis, and B. Ghaleb, \"A Novel Feature-Aware Chaotic Image Encryption Scheme For Data Security and Privacy in IoT and Edge Networks,\" arXiv:2505.00593, 2025. [2] H. Sharma and S. Kaur, “Quantum-inspired hyperchaotic Bio-DNA image encryption for real-time medical security,” 2025. [3] M. M. Elamir et al., “Hybrid image encryption scheme for secure E-health systems,” 2021. [4] M. Maazouz et al., “FPGA implementation of a chaos-based image encryption algorithm,” 2022. [5] Wu Y, Zhang L, Berretti S, Wan S., \"Medical Image Encryption by Content-Aware DNA Computing for Secure Healthcare,\" IEEE Trans. Industrial Informatics, 19(2):2089–2098, 2023. [6] Lai Q., Hu G., Erkan U., Toktas A., \"High-efficiency medical image encryption method based on 2D Logistic-Gaussian hyperchaotic map,\" Applied Mathematics and Computation, 442:127738, 2023. [7] Bentouila et al., \"Secure Medical Image Transmission via CNN-Derived Keys and Chaotic DNA Encoding,\" Indonesian Journal of Electrical Engineering and Informatics (IJEEI), 2025. [8] Kahla M.E.H., Beggas M., Laouid A. et al., \"An IoMT image crypto-system based on spatial watermarking and asymmetric encryption,\" Multimedia Tools and Applications, 83:86681–86706, 2024. [9] S. Banu and R. Amirtharajan, “A robust medical image encryption in dual domain: Chaos-DNA-IWT combined approach,” Medical & Biological Engineering & Computing, 2021. [10] D. Ravichandran et al., “An efficient medical image encryption using hybrid DNA computing and chaos in transform domain,” Medical & Biological Engineering & Computing, 2021. [11] B. Acharya et al., “MIE-SPD: A new and highly efficient chaos-based multiple image encryption technique with synchronous permutation diffusion,” IEEE Access, 2025. [12] W. El-Shafai et al., “Proposed 3D chaos-based medical image cryptosystem for secure cloud-IoMT eHealth communication services,” 2024. [13] D. K. R. Basani et al., “An IoMT-enabled surgical monitoring system utilizing robotics and AI with E2ARiA-RESNET-50 and MI-KMEANS,” 2025. [14] B. Singh, “3-D hyperchaotic quantum-inspired image encryption scheme for secure communication in e-healthcare,” The Journal of Supercomputing, 2025. [15] A. Shafique et al., “A hybrid encryption framework leveraging quantum and classical cryptography for secure transmission of medical images in IoT-based telemedicine networks,” Scientific Reports, 2024. [16] M. Ahmad, “Shuffle medical image encryption scheme based on 4D memristive hyperchaotic map,” Nonlinear Dynamics, 2025. [17] S. Subathra and V. Thanikaiselvan, “Enhanced security for medical images using a new 5D hyper chaotic map and deep learning based segmentation,” Scientific Reports, 2025. [18] B. Yang, “Hyperchaotic color image encryption using eight-base DNA complementary rules and extended Zigzag transform,” PLOS ONE, 2025. [19] Quantum chaotic techniques for medical image encryption in next-generation IoMT applications,” The Journal of Supercomputing, 2025. [20] Odeh A., Abu Al-Haija Q., \"Medical image encryption techniques: a technical survey and potential challenges,\" International Journal of Electrical and Computer Engineering (IJECE), 13(3):3170, 2023. [21] Islam U.O.M., Parah A.S., \"Fast and Lightweight Image Cryptosystem for IoMT Applications,\" Internet of Things, 25:101083, 2024. [22] S. Afzal, M. U. Bokhari, M. Luqman, and B. Hanafi, \"Lightweight CNN-based edge computing with hybrid chaotic encryption for secure and efficient image transmission in IoT networks,\" Discover Computing, vol. 29, art. 65, 2026. [23] M. Zhou, X. Li, W. Du, J. Li, and Z. Wei, “Pixel-level and DNA-level image encryption method based on five-dimensional hyperchaotic system,” Entropy, 2025.

Copyright

Copyright © 2026 Bellapu Lakshmi Amulya. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.