Scholarlink

<p>Multi-medical image encryption faces the significant challenge of balancing image quality with encryption efficiency. This paper introduces a novel time series-oriented algorithm specifically designed for encrypting medical images, particularly brain CT scans. Inspired by the H.264 video coding principles, the proposed algorithm effectively fuses medical image sequences, which enhance the encryption efficiency and the recovery accuracy. Firstly, the multi-CT images are divided into several units, and each unit comprises a key image and the remaining common images. Secondly, in every unit, SURF feature extraction is employed to construct transformation matrices between the key image and all other common images, which are then embedded into the key image using LSB steganography after matrix processing. Finally, block encryption is applied to the fused images using keys derived from SHA-512 and 2D-LPCMM. Pixel scrambling and diffusion are further achieved using replacement rules and row-index tables to obtain the final ciphertext image. Experimental results demonstrate that the proposed algorithm maintains higher fidelity for the decrypted images while maintaining available computation efficiency. Moreover, the proposed encryption algorithm exhibits resistance to various attacks and provides a robust technical solution for the secure transmission and storage of medical images.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Time series-oriented multi-brain CT image encryption with H.264 coding efficiency

  • 期刊论文

  • 发表日期：

    2025年5月29日

查看全文
  • Di Jiang,
  • Xianhua Song,
  • Ying Liu,
  • Ahmed A. Abd El-Latif

摘要

Multi-medical image encryption faces the significant challenge of balancing image quality with encryption efficiency. This paper introduces a novel time series-oriented algorithm specifically designed for encrypting medical images, particularly brain CT scans. Inspired by the H.264 video coding principles, the proposed algorithm effectively fuses medical image sequences, which enhance the encryption efficiency and the recovery accuracy. Firstly, the multi-CT images are divided into several units, and each unit comprises a key image and the remaining common images. Secondly, in every unit, SURF feature extraction is employed to construct transformation matrices between the key image and all other common images, which are then embedded into the key image using LSB steganography after matrix processing. Finally, block encryption is applied to the fused images using keys derived from SHA-512 and 2D-LPCMM. Pixel scrambling and diffusion are further achieved using replacement rules and row-index tables to obtain the final ciphertext image. Experimental results demonstrate that the proposed algorithm maintains higher fidelity for the decrypted images while maintaining available computation efficiency. Moreover, the proposed encryption algorithm exhibits resistance to various attacks and provides a robust technical solution for the secure transmission and storage of medical images.