Reversible watermarking for knowledge digest embedding and reliability control in medical images

Area of Science:

• Medical Imaging
• Digital Watermarking
• Knowledge Representation

Background:

• Medical image sharing for e-learning and remote diagnosis requires enhanced usability and data integrity.
• Current methods may lack efficient retrieval mechanisms based on clinical findings.
• Ensuring authenticity and integrity of medical images during network transmission is crucial.

Purpose of the Study:

• To develop a method for watermarking medical images with a knowledge digest (KD).
• To enable retrieval of similar medical images based on findings or differential diagnoses.
• To embed reliability proofs (integrity and authenticity) within watermarked images.

Main Methods:

• A knowledge digest (KD) is created, summarizing image descriptions, semiology, and similarity rules.
• The KD is embedded into the pixel gray-level values of medical images using watermarking techniques.
• The watermarking approach avoids modifying the image file format.

Main Results:

• Watermarking embeds KD into image pixels, enhancing usability for sharing.
• The KD facilitates retrieval of images with similar findings or differential diagnoses.
• Watermarking provides integrity and authenticity proofs for shared medical images.

Conclusions:

• Watermarking medical images with KDs improves their utility for e-learning and remote diagnosis.
• This technique offers a robust method for secure and searchable medical image databases.
• The approach demonstrates practical application in updating distributed user databases for endoscopic semiology e-learning.