Watermarking Protocol Inspired Kidney Stone Segmentation in IoMT.

Area of Science:

• Medical Imaging and Informatics
• Cybersecurity in Healthcare
• Artificial Intelligence in Medicine

Background:

• Explosion of medical data and smart healthcare demands create significant authentication and integrity challenges.
• Cybercrime targeting healthcare data threatens patient privacy, trust, and diagnostic accuracy.
• Existing systems lack robust solutions for securing medical data in Internet of Medical Things (IoMT) environments.

Purpose of the Study:

• To propose a robust healthcare system integrating kidney stone segmentation with a watermarking protocol for IoMT.
• To enhance authentication, integrity verification, and patient privacy for medical images.
• To improve the accuracy of medical image analysis through advanced segmentation techniques.

Main Methods:

• Generation of chaotic keys from patient data and biometrics for obfuscation and randomization.
• Imperceptible watermark embedding using Singular Value Decomposition (SVD) and adaptive quantization.
• Implementation of a U-Net architecture with a ResNeXt-50 encoder and attention-guided decoder for feature learning.

Main Results:

• Successful integration of a kidney stone segmentation framework with a secure watermarking protocol.
• Demonstrated ability to ensure secure access to unaltered medical data through watermark verification.
• Superior performance compared to state-of-the-art techniques in comprehensive experiments on kidney CT scans.

Conclusions:

• The proposed system offers a robust solution for securing medical data in IoMT applications.
• The integration of watermarking and segmentation enhances both data security and diagnostic capabilities.
• The system provides a practical and effective approach to address challenges in smart healthcare data management.