Scholarlink

Purpose <p>The digital revolution has plunged every facet of human lives. The unrestrained data breaches in the healthcare sector raise the concern about safeguarding the integrity, copyright, and authenticity of patient data. Towards this end, we present an automatic segmentation-based data-hiding method for simultaneous authentication and copyright protection of medical images.</p> Methods <p>We develop an automatic image segmentation algorithm exploiting the morphological characteristics of the image. The Region of Interest (ROI) is extracted using the proposed segmentation method, then the binary mask has been used for implantation of the watermark in ROI and Region of Non-Interest (RONI). ROI is embedded reversibly while NROI is embedded with a robust watermark and patient information using the proposed 2-level DWT algorithm.</p> Results <p>We evaluated our method using the test images from the OPENi medical image repository. The proposed method achieves an average PSNR of 49.61 dB with SSIM of 0.9958 proving its imperceptivity despite a significant payload. Near unity values of NCC for various attacks like filtering, compression and noise addition etc., testify the robustness of our scheme.</p> Conclusion <p>An efficient data hiding scheme for simultaneous authentication and copyright protection of medical images has been proposed. Our scheme has two major advantages, firstly the overhead data for extraction of ROI is not required, and secondly, the reversible data hiding method does not require the look-up table for recovery of the image at the receiver. Experimentation shows that the proposed scheme outperforms state of art terms of visual quality, robustness, payload, and tamper detection capability.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Automatic segmentation based watermarking technique for medical images in e-healthcare

  • 期刊论文

  • 发表日期：

    2026年1月31日

查看全文
  • Solihah Gull,
  • Muzamil Hussan,
  • Shabir A. Parah

摘要

Purpose

The digital revolution has plunged every facet of human lives. The unrestrained data breaches in the healthcare sector raise the concern about safeguarding the integrity, copyright, and authenticity of patient data. Towards this end, we present an automatic segmentation-based data-hiding method for simultaneous authentication and copyright protection of medical images.

Methods

We develop an automatic image segmentation algorithm exploiting the morphological characteristics of the image. The Region of Interest (ROI) is extracted using the proposed segmentation method, then the binary mask has been used for implantation of the watermark in ROI and Region of Non-Interest (RONI). ROI is embedded reversibly while NROI is embedded with a robust watermark and patient information using the proposed 2-level DWT algorithm.

Results

We evaluated our method using the test images from the OPENi medical image repository. The proposed method achieves an average PSNR of 49.61 dB with SSIM of 0.9958 proving its imperceptivity despite a significant payload. Near unity values of NCC for various attacks like filtering, compression and noise addition etc., testify the robustness of our scheme.

Conclusion

An efficient data hiding scheme for simultaneous authentication and copyright protection of medical images has been proposed. Our scheme has two major advantages, firstly the overhead data for extraction of ROI is not required, and secondly, the reversible data hiding method does not require the look-up table for recovery of the image at the receiver. Experimentation shows that the proposed scheme outperforms state of art terms of visual quality, robustness, payload, and tamper detection capability.