Fingerprint-based robust medical image watermarking in hybrid transform

Area of Science:

• Medical Imaging
• Digital Watermarking
• Biometrics

Background:

• Medical image integrity is crucial for diagnosis and treatment.
• Protecting patient privacy and ensuring data authenticity are paramount in healthcare.
• Existing digital watermarking schemes often face challenges in balancing imperceptibility, robustness, and security.

Purpose of the Study:

• To propose a non-blind medical image watermarking scheme using a hybrid transform.
• To enhance the authentication, identification, and privacy of patients through fingerprint watermarking.
• To improve the imperceptibility and robustness of watermarked medical images.

Main Methods:

• A hybrid watermarking scheme combining Lifting Wavelet Transform (LWT) and Discrete Wavelet Transform (DWT).
• Adaptive calculation of scaling and embedding factors using Local Binary Pattern (LBP) values.
• Two-level decomposition: LWT for the first level, DWT for the second.
• Non-blind recovery of the fingerprint watermark.

Main Results:

• The proposed LWT-DWT scheme demonstrates superior imperceptibility and robustness compared to existing methods.
• Effective authentication and identification of original medical images were achieved.
• The scheme showed high resilience against various medical image attacks.

Conclusions:

• The hybrid LWT-DWT watermarking scheme offers a robust and secure solution for medical image protection.
• Patient fingerprint watermarking enhances data security and privacy in medical imaging.
• The adaptive factor calculation contributes significantly to the scheme's performance.