Medical Image Authentication Method Based on the Wavelet Packet and Energy Entropy

Area of Science:

• Digital Image Processing
• Medical Informatics
• Information Security

Background:

• Digital medical information is vulnerable to data compression, noise, scaling, and labeling during transmission.
• Unauthorized copying and malicious tampering pose significant risks to the integrity and copyright of medical data.
• Robust authentication methods are crucial for ensuring the trustworthiness of digital medical information.

Purpose of the Study:

• To design a new method for medical image authentication.
• To ensure the copyright protection and integrity of digital medical information.
• To develop a technique robust against various image processing operations and malicious attacks.

Main Methods:

• A novel medical image authentication method based on wavelet packet and energy entropy is proposed.
• A sliding window approach measures the energy of detailed information within a time-frequency distribution.
• Local energy entropy is constructed, normalized using energy weights, and used as a feature vector for authentication.

Main Results:

• The proposed method successfully mines local details and quantitatively describes energy complexity.
• Experimental results demonstrate strong robustness against a wide range of attacks.
• Attacks tested include shearing, median filtering, contrast/brightness enhancement, various noise types, rotation, scaling, sharpening, and JPEG compression.

Conclusions:

• The developed wavelet packet and energy entropy-based method provides effective authentication for digital medical images.
• The technique offers significant resilience against common signal processing manipulations and security threats.
• This approach contributes to enhanced security and reliability in digital medical information systems.