Image-based decision making for reliable and proper diagnosing in NIFTI format using watermarking.

Area of Science:

• Medical Imaging and Informatics
• Digital Image Processing
• Cybersecurity in Healthcare

Background:

• Advancements in MRI and CT-Scan generate NIFTI images crucial for diagnosing diseases like COVID-19.
• The widespread internet transmission of NIFTI images raises concerns about data integrity, copyright, and accurate diagnosis.
• Existing watermarking schemes lack specific application and robust performance for NIFTI image formats.

Purpose of the Study:

• To develop a secure and robust watermarking scheme specifically for NIFTI neuroimaging data.
• To address the challenges of data integrity, copyright protection, and ensuring diagnostic accuracy of medical images.
• To improve the reliability of NIFTI image communication over networks.

Main Methods:

• A hybrid watermarking approach combining Lifting Wavelet Transform (LWT), Multiresolution Singular Value Decomposition (MSVD), and QR factorization.
• Insertion of multiple watermarks into the initial slice of NIFTI images.
• Evaluation of the scheme's robustness against noise attacks and performance metrics like PSNR, SNR, SSIM, and Normalized Correlation (NC).

Main Results:

• The proposed watermarking scheme demonstrates high image quality (0.99994-0.99998) and SSIM (0.94-0.99).
• Achieved PSNR values range from 56.76 to 57.28 dB, significantly outperforming existing methods (32.66-52.02 dB).
• Normalized Correlation (NC) values between 0.9993 and 0.9998 indicate strong watermark imperceptibility and robustness.

Conclusions:

• The hybrid LWT, MSVD, and QR-based watermarking scheme offers superior robustness and security for NIFTI images.
• This method effectively protects the integrity and copyright of medical images, crucial for reliable healthcare diagnostics.
• The proposed technique provides a significant advancement in securing neuroimaging data transmission and storage.