Medical Image Watermarking for Secure E‐Health Application

Abstract

The rapid advancement of communication technologies creates significant chal- lenges in protecting medical imaging data. Digital watermarking has emerged as a

formidable solution, embedding inconspicuous yet traceable information into mul- timedia content to safeguard data security and authenticity, especially in health- care. Unlike other methods, watermarking intrinsically links the information to

the data, providing robust protection against tampering and unauthorized modi- fications.

In e-healthcare, the secure transfer of medical images is crucial for services such as telemedicine and teleradiology, as it improves diagnostic accuracy and protects patient data. This study introduces advanced blind watermarking methods based on biometric technology to ensure secure transmission while maintaining image integrity and authenticity. The proposed techniques are designed to balance the practical challenges of detection rates, watermark visibility, and overall robustness. This work presents robust watermarking frameworks to enhance medical image

management in e-healthcare systems, highlighting practical benefits and chal- lenges. It integrates advanced security methods like encryption, blockchain, and

compression algorithms, which will be tested in realistic scenarios to validate their effectiveness. By combining these innovative solutions, this research supports the creation of secure and efficient systems for electronic medical records.

This thesis introduces two advanced watermarking frameworks founded on bio- metric technology to ensure secure image transmission while preserving integrity.

The first, a Multi-Layered Security Framework (MLSF), integrates compressed fingerprint watermarking with encryption and blockchain. The second approach, Improved DWT and ACM Watermarking (IDAW), combines palmprint features with chaotic transformations to enhance the watermark’s resilience. Experimental evaluations confirm the effectiveness of these contributions. The MLSF demonstrates exceptional imperceptibility, achieving a PSNR of 63.24 dB and an SSIM of 1.0. The IDAW framework proves highly robust against a range of attacks, including compression and noise, maintaining a PSNR of 53.95 dB and an

SSIM of 0.99996. By integrating innovative security solutions, this research con- tributes to the development of secure and efficient systems for managing electronic

medical records in the modern digital era.