Embedding Biometric Information in Interpolated Medical Images with a Reversible and Adaptive Strategy.

Area of Science:

• Cybersecurity
• Digital Image Processing
• Medical Informatics

Background:

• Reversible data hiding (RDH) is crucial for secure digital image applications, particularly in medical image management.
• Embedding patient biometric data into medical images aids authentication and ensures data integrity.
• Existing RDH schemes face challenges in balancing embedding capacity and image distortion.

Purpose of the Study:

• To develop an adaptive interpolation-based reversible data hiding (RDH) scheme for secure embedding of biometric information into medical images.
• To enhance data security and authentication in medical image management systems.
• To improve upon existing RDH techniques by minimizing image distortion while maximizing data payload.

Main Methods:

• The proposed scheme classifies image pixel blocks into complex and smooth categories based on pixel value dynamics.
• Complex blocks utilize minimum-neighbor (MN) interpolation and Difference-Maximization (DIM) embedding.
• Smooth blocks employ block mean (BM) interpolation, prediction error histogram (PEH) embedding, and difference expansion (DE) embedding.

Main Results:

• The adaptive strategy achieves low distortion for smooth blocks and a high payload for complex blocks.
• Experimental results demonstrate the scheme's effectiveness for both medical and general images.
• The proposed scheme shows a peak signal-to-noise ratio (PSNR) improvement of up to 10.32 dB over state-of-the-art methods.

Conclusions:

• The developed adaptive RDH scheme offers a robust solution for embedding sensitive data in digital images, especially for medical applications.
• The method provides superior performance in terms of image quality and data embedding capacity compared to previous approaches.
• This technique enhances the security and reliability of medical image management and authentication systems.