Abstract
Internet of Medical Things (IoMT)-driven smart health and emotional care is revolutionizing the healthcare industry by embracing several technologies related to multimodal physiological data collection, communication, intelligent automation, and efficient manufacturing. The authentication and secure exchange of electronic health records (EHRs), comprising of patient data collected using wearable sensors and laboratory investigations, is of paramount importance. In this article, we present a novel high payload and reversible EHR embedding framework to secure the patient information successfully and authenticate the received content. The proposed approach is based on novel left data mapping (LDM), pixel repetition method (PRM), RC4 encryption, and checksum computation. The input image is upscaled by using PRM that guarantees reversibility with lesser computational complexity. The binary secret data are encrypted using the RC4 encryption algorithm and then the encrypted data are grouped into 3-bit chunks and converted into decimal equivalents. Before embedding, these decimal digits are encoded by LDM. To embed the shifted data, the cover image is divided into $2 \times 2$ blocks and then in each block, two digits are embedded into the counter diagonal pixels. For tamper detection and localization, a checksum digit computed from the block is embedded into one of the main diagonal pixels. The average peak signal to noise ratio (PSNR) of the stego-images obtained is 41.95 dB for a very high embedding capacity of 2.25 bits per pixel. Furthermore, the embedding time is less than 0.2 s. Experimental results reveal that our approach outperforms many state-of-the-art techniques in terms of payload, imperceptibility, computational complexity, and capability to detect and localize tamper.
The Challenge of Secure IoMT
As healthcare shifts toward "smart health," patient data is increasingly collected via wearable sensors and shared across edge networks. This creates a massive security challenge: how do you transmit sensitive Electronic Health Records (EHR) alongside medical imagery (like X-rays or CT scans) without risking data breaches or compromising the image quality?
In medical imaging, "reversibility" is non-negotiable. Any permanent distortion introduced by data hiding could lead to a misdiagnosis. Furthermore, edge devices (like tablets or local gateways) lack the computational power for heavy cryptographic or transform-domain operations. The authors aim to fill the gap for a system that is simultaneously high-capacity, computationally lightweight, reversible, and capable of tamper detection.
The Proposed Framework
The proposed framework, a type of Reversible Information Hiding (RIH), relies on four main pillars to achieve its performance goals.
- Pixel Repetition Method (PRM): Instead of complex interpolation, the system upscales an input image by simply repeating pixels to create $2 \times 2$ blocks. This provides the "space" to hide data while ensuring the original "seed" pixels remain untouched and easily recoverable.
- RC4 Encryption: The secret EHR data and a fragile logo (for authentication) are encrypted using the RC4 stream cipher before embedding, providing a baseline layer of data security.
- Left Data Mapping (LDM): This is the paper's primary innovation. Standard data hiding often shifts pixel values significantly, creating visual noise. LDM maps 3-bit data chunks (0–7) to a signed range (-4 to 3). By shifting the data toward the negative axis of the number line, the average modification to each pixel is reduced, significantly improving the image's "imperceptibility" (PSNR).
- Block-Based Authentication: The $2 \times 2$ blocks are used strategically. Two pixels hold the encrypted EHR data, while a third pixel stores a checksum digit. This allows the receiver to verify the integrity of the data at a very granular level.
Key Insights and Ideas
- The "Aha" Moment of LDM: The researchers realized that the visual impact of data hiding isn't just about how much data you hide, but the magnitude of the change to the pixel value. By remapping the data range using LDM, they reduced the average pixel difference from 3.5 units to just 2.0 units, making the hidden data much harder to detect with the human eye.
- Spatial Domain Efficiency: While many researchers use transform domains (like Discrete Wavelet Transform), this paper stays in the spatial (pixel) domain. This avoids the heavy math required for transforms, making it fast enough for real-time use on low-power edge hardware.
- Fragility as a Feature: The system uses a "fragile" watermark. While "robust" watermarks are designed to survive image editing, a fragile one is designed to break if even a single pixel is altered. In a medical context, this alerts the doctor immediately if the record has been tampered with during transit.
Key Results
The authors evaluated their framework using medical images (X-rays, MRI, CT), the UCID database, and standard test images. The results demonstrate a significant improvement over existing methods.
- High Payload: The system achieves an embedding capacity of 2.25 bits per pixel (bpp), outperforming several recent methods that hover around 0.5 to 1.5 bpp.
- Superior Quality: Despite the high payload, the average Peak Signal-to-Noise Ratio (PSNR) remained at 41.95 dB. Generally, any value over 35 dB is considered invisible to the human eye.
- Speed: The embedding time is remarkably low—less than 0.2 seconds per image on average, which is roughly 3x faster than the benchmarks cited in the paper.
- Tamper Localization: The system successfully detected and localized "copy-paste" attacks, showing exactly which $2 \times 2$ blocks were modified.
Conclusion
The proposed scheme offers a robust balance of security, capacity, and speed, making it a potential candidate for providing better security and authentication solutions for IoMT-based smart health. Future work may involve exploring more modern lightweight ciphers and self-healing mechanisms to restore tampered regions.
Summary updated