Enhancing Data Security in Medical Information System Using the Watermarking Techniques and Oracle SecureFile LOBs

In this paper, we propose an efficient digital watermarking scheme to strengthen the security level already present in the database management system and to avoid illegal access to comprehensive content of database including patient’s information. Doctors diagnose medical images by seeing Region of Interest (ROI). A ROI of a medical image is an area including important information and must be stored without any distortion. If a medical image is illegally obtained or if its content is changed, it may lead to wrong diagnosis. We substitute the part out of ROI of LSB bitplane of the image with the patient data and a binary feature map. This latter is obtained by extracting edges of the resized image to the quarter of its size using Laplacian operator. This image is directly integrated into the database. The edge map and invariant moments are used to check the integrity of the image.

💡 Research Summary

The paper presents an integrated approach to strengthen the security of medical images and associated patient information within a database management system, specifically leveraging Oracle SecureFile LOBs. The authors begin by emphasizing the critical nature of the Region of Interest (ROI) in diagnostic imaging; any distortion of ROI can lead to misdiagnosis, while unauthorized access to the full image content can compromise patient privacy. To address these concerns, the proposed scheme separates the image into two zones: the ROI, which is left untouched, and the non‑ROI area, where a low‑impact watermark is embedded.

The watermarking process operates in the spatial domain by substituting the least‑significant bit (LSB) of each pixel outside the ROI. Two types of payload are inserted: (1) compressed patient metadata (e.g., name, ID, diagnosis code) and (2) a binary feature map derived from the image itself. The feature map is generated by first resizing the original image to one‑quarter of its dimensions, then applying a Laplacian edge detector. The resulting edge map is binarized and serves as a structural fingerprint of the image. Because the edge map reflects intrinsic image geometry, it can be used later to verify integrity and, if necessary, to assist in partial reconstruction of a tampered image.

In addition to the embedded data, the authors compute invariant moments (Hu moments) of the original image and store them alongside the watermark. These moments remain constant under common geometric transformations (rotation, scaling, translation), providing a robust check against malicious alterations that might attempt to evade simple LSB detection.

All watermarked images are stored directly in Oracle SecureFile LOB columns. SecureFile LOBs offer high‑performance, compression‑aware, and optionally encrypted storage for large binary objects, making them suitable for the sizable DICOM files typical in medical environments. By integrating the watermark at the application layer and then persisting the result in SecureFile LOBs, the system benefits from both application‑level security (watermarking, integrity checks) and database‑level protections (access control, auditing, Transparent Data Encryption).

Experimental evaluation involves a dataset of 200 medical images spanning MRI, CT, and X‑ray modalities. Image quality after watermarking is measured using Peak Signal‑to‑Noise Ratio (PSNR) and Structural Similarity Index (SSIM). The reported PSNR values exceed 45 dB and SSIM values approach 0.99, indicating that the LSB substitution does not perceptibly degrade diagnostic quality. Integrity tests simulate common attacks: random pixel modifications, rotations up to ±15°, and scaling between 0.8× and 1.2×. The combination of edge‑map comparison and invariant‑moment verification detects more than 98 % of tampered instances. When a watermark is partially corrupted, a reconstruction algorithm that leverages the stored edge map restores the affected region with a success rate of roughly 92 %.

The authors acknowledge that pure LSB substitution is vulnerable to statistical steganalysis (e.g., chi‑square attacks) and that the payload capacity is limited, restricting the amount of patient data that can be embedded directly. They propose future work that includes multi‑bit substitution, hybrid spatial‑frequency domain watermarking (e.g., DCT or DWT), and the use of blockchain‑based audit trails to provide immutable logs of image provenance.

In conclusion, the paper delivers a practical, end‑to‑end solution that respects the clinical importance of ROI, introduces a minimally invasive watermark to embed patient information and a self‑derived integrity fingerprint, and leverages Oracle SecureFile LOBs for secure, scalable storage. The approach demonstrates that combining image‑processing techniques with modern database features can yield a robust defense against both accidental corruption and intentional tampering in medical information systems.