Combining I-BSIF Fingerprints and FlexenTech for Reversible Medical Image Watermarking

The rise of digital medical imaging in telemedicine raises concerns about image authenticity and patient privacy. Traditional watermarking methods struggle to achieve robust security and reliable patient authentication simultaneously. This study presents a novel blind reversible watermarking framework that combines orientation-guided Improved Binary Statistical Image Features (I-BSIF) with FlexenTech encryption to improve medical image security. The method utilizes advanced fingerprint feature extraction tailored to local ridge orientations, resulting in compact biometric signatures. These features undergo lightweight FlexenTech encryption before being embedded into medical images through spread-spectrum modulation in mid-frequency subbands of the Lifting Wavelet Transform. Experimental results demonstrate that rotation-invariant feature extraction preserves more than 99% similarity after a 270° rotation, with an average Peak Signal-to-Noise Ratio (PSNR) exceeding 51 dB and a Structural Similarity Index (SSIM) of nearly 0.999, confirming the method's effectiveness in maintaining patient authentication and image integrity.

• reversible watermarking
• medical image watermarking
• blind watermarking
• spread spectrum
• lifting wavelet transform
• rotation invariance
• image integrity
• patient authentication
• image authenticity
• telemedicine security

• Advanced Steganography and Watermarking Techniques
• Biometric Identification and Security
• Chaos-based Image/Signal Encryption
• reversible watermarking
• medical image watermarking
• blind watermarking
• spread spectrum
• lifting wavelet transform
• rotation invariance
• image integrity
• patient authentication
• Computer Vision and Pattern Recognition
• Signal Processing
• Computer Science