A secure framework for remote diagnosis in health care: A high capacity reversible data hiding technique for medical images
Graphical abstract
Introduction
In the digital era, the protection of confidential data from unauthorized access is highly demanded. Reversible data hiding (RDH) plays an integral role in secure communication because the embedding process is done in such a way that the hidden data is unnoticeable to attackers. It has a variety of applications in the area of medical imagery, military imagery, etc. Clinical information can be embedded in a medical image and transmitted to a doctor who is sitting anywhere across the globe. Even a minute alteration in the original medical image is unacceptable for correct diagnosis. One can use the data hiding algorithm, which reversible [1, 2], so that the receiver can extricate the hidden data and also construct the original image.
Usually, the two main classifications of data hiding algorithms are reversible and irreversible. In the irreversible category, the restoration of the original image is not possible, but it renders high embedding rate. The effectiveness of a data-hiding scheme is evaluated mainly using parameters like embedding capacity and visual quality. Embedding capacity or payload capacity gives the total number of secret bits, which can be embedded into a cover image using the given algorithm. Succeeding the data embedding process, the cover image is called the stego image and is needed to compute the visual similarity between these two images. Normally, there is a trade-off between these two parameters. Thus, it is a challenging research area. Recently, a substantial bulk of work has been done in the field of RDH. Some researchers developed an algorithm in the spatial domain, while some have utilized frequency domain.
Image-based RDH schemes are appropriate for applications like medical imagery because we can hide large amounts of electronic patient record (EPR) data into a medical image. Interpolation is a technique used to find out unknown values using known information. Interpolation is used for up-sampling and the interpolated pixels are used for data embedding. A few research works are based on this idea and the original pixels are untouched in the process of data embedding. A novel approach has been put into use in this paper where along with the interpolated pixels, the original pixels were also utilized for data embedding so that the payload capacity of the algorithm is very high and we could reconstruct the original image as such. Recently remote diagnosis plays a key role in health care [3], and the security aspect of medical images in this context has gained significance.
The main stimulus for this work is the desire to implement a high-capacity RDH scheme suitable for medical applications. The considerable offerings of the proposed work are:
- 1
A new image interpolation technique that outperforms the recently published interpolation techniques.
- 2
Implementation of a high-capacity data-hiding algorithm, which is reversible in nature.
- 3
On the basis of the analysis of the latest literature in accordance to the up-sampling and data embedding, it was observed that the original pixels were untouched. To boost the payload capacity of the proposed algorithm, the original pixels have also been utilized.
- 4
The outcome of the experiments authenticate that the prospective algorithm produces excellent visual quality stego images.
The remaining paper is structured in this way; Sect. 2 examines the analogous work in this area. Sect. 3 discusses the recommended method, and experimental results and analysis are given in Section 4. Finally, the concluding remarks are given.
Section snippets
Related works
Various internet of things (IoT) based electronic health architectures have been proposed but securing EPR data still needs research attention. Tian [4] developed a difference expansion technique for data hiding, and their maximum embedding capacity is 0.5 bit per pixel (bpp) only. Lou et al. [5] suggested an RDH scheme, which was based on adaptive difference expansion and their results are better than the previous methods. Pawar et al. [6] suggested a method based on shifting of the histograms
Proposed method
Fig. 1 exemplifies the block diagram of the propounded algorithm. Initially, the original input image is up-sampled using the proposed interpolation technique discussed in subSection 3.1 to produce the cover image. After the cover image generation, the data embedding process is performed. Secret data is embedded into the interpolated pixel using a modular arithmetic operation discussed in subSection 3.2. To boost the payload capacity of the algorithm original reference pixels are also utilized
Experimental results and analysis
To prove the efficiency of the proposed technique, experiments were carried out on nine standard images (shown in Fig. 4) of the general category taken from literature. The technique was then applied to six medical images (shown in Fig. 5) obtained from the DICOM library to demonstrate how efficiently the electronic health records can be embedded in the image. For embedding purposes, we have used data generated by a pseudo-random function. Experiments were conducted using MATLAB R2015a.
Eq. (18)
Conclusions
A high-capacity RDH technique mastered for disclosing clinical information via medical images has been proposed. The developed scheme exploits the concept of image interpolation. A new weighted interpolation has been designed which assigns additional weightage to the nearest pixels and our proposed image interpolation dominates few interpolation techniques of recent times. Also, the interpolation in the most recent works consumed more computational time for generating the cover image because it
CRediT authorship contribution statement
P.V. Sabeen Govind: Conceptualization, Methodology, Data curation, Writing - original draft. M.V. Judy: Supervision, Validation, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Sabeen Govind P V is currently pursuing Ph.D. from Cochin University of Science and Technology, Kerala, India. He completed his M. Tech in Computer Science from University of Kerala. He is working as an Assistant Professor in Rajagiri College of Social Sciences (Autonomous), Kerala. His research interests include Image processing and Data hiding.
References (25)
- et al.
An efficient image interpolation increasing payload in reversible data hiding
Expert Syst Applica, Elsevier
(2012) - et al.
Hiding clinical information in medical images: a new high capacity and reversible data hiding technique
J Biomed Inform, Elsevier
(2017) - et al.
Data hiding method using image interpolation
Comput Stand Interfaces, Elsevier
(2009) - et al.
Electronic health record hiding in images for smart city applications: a computationally efficient and reversible information hiding technique for secure communication
Futu Gener Comput Syst, Elsevier
(2020) - et al.
Reversible dynamic secure steganography for medical image using graph coloring
Health Policy Technol, Elsevier
(2013) - et al.
Information hiding scheme for digital images using difference expansion and modulus function
J King Saud Univ Comput Inform Sci, Elsevier
(2019) - et al.
Enhanced AMBTC based data hiding method using hamming distance and pixel value differencing
J Inform Secur Applica, Elsevier
(2019) - et al.
A reversible and secure patient information hiding system for IoT driven e-health
Int J Inform Manag, Elsevier
(2019) - et al.
Hiding Electronic Patient Record (EPR) in medical images: a high capacity and computationally efficient technique for e-healthcare applications
J Biomed Inform, Elsevier
(2017) - et al.
Reversible steganography using extended image interpolation technique
Comput Electri Eng, Elsevier
(2015)
High capacity reversible hiding scheme based on interpolation, difference expansion and histogram shifting
Multimed Tools Appl, Springer
Reversible data embedding using a difference expansion
IEEE Trans Circu Syst Video Technol
Cited by (12)
Separable and reversible data hiding scheme for medical images using modified Logistic and interpolation
2024, Biomedical Signal Processing and ControlSteganalysis of adaptive multi-rate speech with unknown embedding rates using clustering and ensemble learning
2023, Computers and Electrical EngineeringReversible data hiding with enhancing contrast and preserving brightness in medical image
2022, Journal of Information Security and ApplicationsCitation Excerpt :And these methods are superior when large amounts of patient-identifying information need to be embedded. Govind et al. [17] proposed a high capacity reversible data hiding technique that can be used to embed patient data using a new weighted interpolation technique, the payload capacity has been improved in the method. Bhardwaj et al. [18] proposed a novel encrypted dual image reversible data hiding, this method gave very high embedding rates on both natural and medical image.
A high-capacity and reversible patient data hiding scheme for telemedicine
2022, Biomedical Signal Processing and ControlCitation Excerpt :But the cover image generated via this method has an identical weakness (i.e. block effect) with the nearest neighbor interpolated image. Afterward, a novel image scaled-up scheme [29] is developed to form a cover image, and modular arithmetic and difference expansion is applied for data embedding. This algorithm achieves an embedding rate of 2.3 bit per pixel (bpp) for peak signal noise ratio (PSNR) of 36 dB.
Advances in medical image watermarking: a state of the art review
2024, Multimedia Tools and Applications
Sabeen Govind P V is currently pursuing Ph.D. from Cochin University of Science and Technology, Kerala, India. He completed his M. Tech in Computer Science from University of Kerala. He is working as an Assistant Professor in Rajagiri College of Social Sciences (Autonomous), Kerala. His research interests include Image processing and Data hiding.
Judy M V, Ph.D. in Computer Science, and currently working as an Associate Professor in the Department of Computer Applications, Cochin University of Science and Technology, Kerala, India. Her main research interests include Computational biology, Evolutionary algorithms, Data mining, Big data analytics and Image processing.
This paper is for CAEE special section SI-ipsa Reviews processed and recommended for publication to the Editor-in-Chief by Area Editor Dr. E. Cabal-Yepez.