IoMT Privacy Preservation: A Hash-Based DCIWT Approach for Detecting Tampering in Medical Data

Smart Healthcare has brought advantages to both healthcare professionals and individuals living in remote areas. However, deliberately or inadvertently altering medical records can result in an inaccurate diagnosis. As a result of telemedicine, patients can easily communicate with their healthcare providers, and the Internet of Medical Things (IoMT) devices enable remote consultations, allowing for timely diagnosis and treatment. Medical data can be protected by watermarking, a combination of steganography, and fingerprinting. This paper proposes a hash-based approach using the Discrete Cosine Integer Wavelet Transform (DCIWT) to provide secure watermarking for confidential medical data. To preserve critical Region of Interest (RoI) integrity, we use SHA-256 cryptographic hash algorithm. The hashed Region of Interest (hRoI) is embedded in the Region of Non-Interest (RoNI) by DCIWT. To enhance the segmentation process, we employ adaptive thresholding, which dynamically adjusts the threshold based on local image characteristics, ensuring accurate segmentation of RoI and RoNI across varying image qualities. The hidden data is significantly affected if the watermarked image is altered. As a result of comparing the original RoI with the extracted RoI, we determine if an image has been altered. Tampered regions are be recovered by extracting the RoI and using it for diagnosis. The proposed digital watermarking technique maintains RoI integrity using message digests and histograms. Through extensive experiment results, we demonstrate that the perceptual properties of the images are intact in terms of Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM).