Related Experiment Video
Updated: Jul 14, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.8K
Blind video watermarking scheme for medical video authentication
Doaa Sami Khafaga1, Manar Alohaly2, Mostafa M Abdel-Aziz3
1Department of Computer Sciences, College of Computer and Information Sciences, Princess Nourah bint Abdulrahman University, P.O. Box 84428, Riyadh 11671, Saudi Arabia.
Heliyon
|October 9, 2023
Summary
This study introduces a novel blind watermarking technique for medical videos, enhancing data security and integrity. The method ensures high confidentiality and robustness against attacks with minimal distortion, improving patient data protection.
Area of Science:
- Medical imaging and digital security
- Information security in healthcare
- Digital watermarking techniques
Background:
- Medical videos contain sensitive patient data requiring robust protection against unauthorized access and redistribution.
- Existing watermarking schemes often struggle to balance security, imperceptibility, and robustness.
- Ensuring data integrity and authenticity is paramount in medical video archiving and transmission.
Purpose of the Study:
- To propose a novel blind watermarking scheme for medical videos.
- To enhance confidentiality, integrity, authenticity, and perceptual quality of medical videos.
- To achieve minimal distortion while maintaining robustness against common attacks.
Main Methods:
- Utilized two-dimensional Discrete Wavelet Transform (2D-DWT) for frame processing and sub-band transformation.
- Applied dual Hessenberg-QR decomposition on the selected HL2 wavelet sub-band for watermark embedding.
- Employed Arnold cat map for watermark scrambling to enhance confidentiality.
- Implemented a fully blind watermark extraction process, eliminating the need for the original video.
Main Results:
- Achieved high perceptual quality with minimal distortion, evidenced by a high PSNR value of 70.6899 dB.
- Demonstrated significant robustness against various attacks, with an average Normalized Correlation (NC) of 0.9998 and Bit Error Rate (BER) of 0.0023.
- Maintained a favorable trade-off between robustness and visual imperceptibility compared to existing methods.
- Showcased superior performance and large payload capacity over similar video watermarking techniques.
Conclusions:
- The proposed blind watermarking scheme effectively secures medical videos, improving confidentiality, integrity, and authenticity.
- The method offers a robust solution with high visual imperceptibility, suitable for sensitive medical data.
- Future work could explore faster decomposition methods to mitigate embedding time, such as bound-constrained SVM.
