Content uploaded by Ammar Odeh
Author content
All content in this area was uploaded by Ammar Odeh on Dec 30, 2023
Content may be subject to copyright.
164
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh
1
* and Anas Abu Taleb2
1* Assistant Professor, Princess Sumaya University for Technology, King Hussein School of
Computing Sciences, Computer Science Department, Jordan. A.odeh@psut.edu.jo,
Orcid: https://orcid.org/0000-0002-9929-2116
2 Associate Professor, Princess Sumaya University for Technology. King Hussein School of
Computing Sciences, Computer Science Department, Jordan. a.abutaleb@psut.edu.jo,
Orcid: https://orcid.org/0000-0002-8286-1829
Received: August 26, 2023; Accepted: October 28, 2023; Published: December 30, 2023
Abstract
In modern healthcare, transmitting digital medical images through open-source networks for
diagnosis in remote centers poses a significant security challenge due to the sensitive patient
information involved. This paper presents an algorithm designed to encrypt these crucial medical
images comprehensively. The algorithm generates hash code to validate image integrity, followed
by feature-based watermarking, transformation into the frequency domain via Discrete Cosine
Transform (DCT), and encryption using Advanced Encryption Standard (AES) and RSA encryption
techniques. Each step plays a pivotal role in fortifying the images against unauthorized access,
tampering, or interception during transmission, Guaranteeing the secrecy, unaltered state, and
legitimacy of the included medical information. Implementing this encryption algorithm mandates
strict adherence to cryptographic best practices, robust execution of encryption algorithms, secure
key management, and compliance with industry standards. These meticulous measures bolster the
confidentiality and integrity of medical images, which are crucial in protecting patient privacy and
maintaining data integrity within healthcare systems. This comprehensive encryption strategy
addresses the need for the purpose of ensuring safe and protected delivery and protection of sensitive
medical information across networks.
Keywords: Homomorphic Encryption, Medical Images, Peak-Signal-to-Noise Ratio, Number of
Pixel Change Rate, Unified Average Changing Intensity, Entropy.
1 Introduction
The HITECH Act enacted in 2009 was a significant catalyst in encouraging the widespread adoption of
Electronic Health Records (EHR), aiming to revolutionize the accessibility and management of patient
data (Gold, M., 2016). EHR systems simplify updating medical information electronically and foster
seamless communication between healthcare providers, marking a monumental shift from traditional
paper-based records (Furukawa, M.F., 2014).
EHR's pivotal role in enhancing healthcare quality cannot be overstated. Its digitized format
consolidates a patient's comprehensive medical history, presenting a myriad of advantages when
Journal of Wireless Mobile Networks, Ubiquitous Computing, and Dependable Applications (JoWUA),
volume: 14, number: 4 (December), pp. 164-176. DOI: 10.58346/JOWUA.2023.I4.012
*Corresponding author: Assistant Professor, Princess Sumaya University for Technology, King Hussein
School of Computing Sciences, Computer Science Department, Jordan.
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
165
compared to conventional paper records. Beyond the sheer capacity to manage vast volumes of patient
data efficiently, EHR systems significantly enhance the overall quality of patient care through swift and
precise record updates (Lite, S., 2020) (Reisman, M., 2017).
In tandem with the increasing reliance on medical imaging for diagnoses and treatment, concerns
about the confidentiality of these images have surfaced. While instrumental in medical advancements,
technologies like MRI and CT scans contain sensitive patient information (Panayides, A.S., 2020). The
leakage or compromise of such data could have severe privacy implications for patients and legal
ramifications for healthcare institutions. Consequently, extensive efforts have been undertaken to
develop robust security solutions, such as cryptographic methods, to safeguard these invaluable medical
images and protect patient privacy (Ramzan, M., 2022) (Kaissis, G.A., 2020).
Securing sensitive information within the digital domain of medical imaging presents a significant
challenge (Rasool, R.U., 2022) (Nosrati, S., 2020). The exposure of these images on shared digital
networks makes them susceptible to a spectrum of potential attacks, necessitating advanced
cryptographic solutions (Singh, S., 2021). It is imperative to employ encryption techniques that render
the data indecipherable, ensuring its confidentiality even if breaches occur or unauthorized access is
gained. Among the various encryption approaches, symmetric encryption stands out for its efficacy in
securing large volumes of data and upholding confidentiality (Shukla, S., 2022) (Bandari, V., 2023).
Innovatively, chaos theory has emerged as a promising frontier in modern cryptography, posing a
challenge to traditional symmetric encryption systems such as AES (Zhang, J., 2020). Chaos-based
cryptographic systems offer many advantages, including sensitivity to initial conditions, deterministic
randomisation, structural complexity, a broad critical space, adaptability, and extensive periodicity.
Developing improved algorithms and implementations in cryptographic solutions remains crucial to
fortify medical images against evolving threats and maintain the utmost confidentiality of sensitive data
in healthcare settings (Teh, J.S., 2020) (El-Latif, A.A.A., 2022).
Moreover, the implementation of such sophisticated encryption techniques not only ensures
confidentiality but also strengthens the integrity of health records. However, the ever-evolving landscape
of cybersecurity demands continuous refinement and vigilance in safeguarding patient information
(Aram, S., 2016). As technology advances, efforts to mitigate potential vulnerabilities in cloud storage
and encryption algorithms remain crucial. Additionally, expanding the application of these secure
methodologies to various aspects of healthcare, such as telemedicine and remote patient monitoring, is
pivotal for ensuring a comprehensive and robust healthcare data security framework.
The Following are the Key Contributions
• Design of improved multi-layered encryption process involving hash code generation, feature-
based watermarking, transformation into the frequency domain using DCT, AES encryption for
data security, and additional security layers through RSA encryption.
• Introduce `critical cryptographic techniques like hashing (SHA-256) to create a unique digital
fingerprint, watermarking to embed the hash information imperceptibly, and frequency domain
transformation (DCT) for enhanced representation of image content, thereby bolstering the
image's resilience against potential attacks and un unauthorised modifications.
• Conduct an assessment to evaluate the execution performance and compare the findings with
recent studies or works for comparative analysis.
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
166
2 Related Work
Numerous methods are commonly employed to safeguard data privacy in cloud storage. One such
prominent technique is the Electronic Health Record (EHR), utilized for digitally managing patients'
healthcare details. Cloud storage is preferred given the vast amount of data involved, although it raises
significant security concerns in cloud computing (Jouini, M., 2019). As a result, numerous cryptographic
algorithms have been suggested to bolster data security within the cloud. Introducing a lightweight
encryption algorithm, the LSFE method has been implemented to provide medical records in a secure,
encrypted format (Yao, F., 2021) (Hamza, N.A., 2019). This algorithm ensures encrypted data; further,
image segmentation is executed for enhanced security. The reduction of images is achieved using
Hadoop and MapReduce. Consequently, this method assures the confidentiality of healthcare data
(Hamza, N.A., 2019).
In a prior study (Dzwonkowski, M., 2015), a method utilizing quaternions to encrypt DICOM images
was presented and juxtaposed against the Advanced Encryption Standard Electronic Codebook (AES-
ECB). This particular technique exhibited significantly swifter performance in contrast to ECB.
Furthermore, in (Wadi, S.M., 2014) introduced an novel encryption method for new-definition medical
image security, employing an adapted version of AES. However, this approach underwent cryptanalysis
in (Yap, W.S., 2016) through an impossible variance cyberattack in subsequent investigations.
In (Li, L., 2012) presented an encrypted algorithm for secret medical images employed the Elliptic
Curve ElGamal encryption algorithm. They selected elliptic curve parameters to resist attacks such as
Pohlig-Hellman, Pollard’s rho (Pollard, J.M., 1978), and Isomorphism. The algorithm involved
converting multiple pixel values into binary, combining them to form large integers below a prime
modulo, and encoding these integers into elliptic curve coordinates utilize the Koblitz encoding
technique. Compared to RSA and ElGamal schemes, Li et al.'s approach exhibited superior execution
speed.
In (Tawalbeh, L.A., 2013) introduced a method for encrypting images content by utilizing elliptic
curve cryptography, introducing two distinct models for medical image Encipherment. One algorithm
involved image compression, followed by encrypting the DC components using ECC. In the second
algorithm, ECC encryption targeted only the most significant bits. Despite ECC's high-security features
relying the way the author has set up their elliptic curve parameter for ECDLP might not be large enough
to effectively guard against attacks like Baby-step, giant-step, or Pollard’s Rho attack. Despite the strong
security features of ECC based on ECDLP, the size of the cyclic order in the elliptic curve parameter
might not be sufficient to fend off assaults such as Baby-step, giant-step, or Pollard’s Rho attack.
In a prior publication (Behnia, S., 2013), a system for encrypting medical images was proposed,
employing Jacobian elliptic maps. The method involved utilizing the Jacobian elliptic map to create a
disorderly sequence, subsequently generating a coded image through XOR operations between this
disordered sequence and the pixel values of the original image. Furthermore, Manish et al. (Kumar, M.,
2016) presented an alternative image encryption strategy in which pixel values were encoded using a
DNA encoding method, followed by circular shifting using keys exchanged through an asymmetric
encryption technique rooted in the Elliptic Curve Diffie-Hellman algorithm
The approach in (Yu, F., 2021) Chaos was employed to enhance security in encrypting two images
simultaneously. The method included encrypting the extracted face from the image and then conducting
a second encryption on the whole picture. The keystream used for the cryptographic system was
produced by the 2D SFSIMM (a Two-dimensional hyperchaotic map, sinusoidal feedback Sine ICMIC
modulation map), blending both scrambling and diffusion techniques. This approach showed robustness
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
167
against diverse statistical attacks because breaking the encryption algorithm necessitated overcoming
two encryption rounds.
In a different study (Yu, F., 2021), A new cryptographic system was created to encode and decode
images using three secure maps. The sine map was used to rearrange pixel coordinates within the original
image through a permutation method, followed by substitution using a second secret key, K. Finally,
scrambling of the image based on CTM (Cipher-based Technique for Mapping) was performed using a
bit XOR operation.
Another proposed technique for chaotic image encryption, outlined in (Cheng, Z., 2022), integrated
Latin squares and random shifts. The algorithm involved key creation, pixel scrambling, pixel swapping,
and bit scrambling. Improving the complexity of the resulting Latin square matrix enhanced the
procedure’s security and robustness. The key was initially generated from the plain image to heighten
sensitivity in the cryptography approach. Pixel position scrambling followed, cyclically shifting each
pixel to the right within every row of the image matrix. Coordinate elements of the image matrix were
then replaced with the lookup table's values from a 256-by-256 Latin square matrix containing a chaotic
sequence, determined based on the image's pixel values and the sequence's values. Implementing
multimedia encryption techniques demands more resources (time and storage). Consequently,
lightweight image encryption techniques, which require minimal memory, time, or energy while
providing high security for low-powered devices, are gaining popularity.
Subsequently, Ferdush et al. (2021) explored lightweight image encryption based on chaos,
introducing a standardised approach and method using two distinct chaotic maps—Arnold and Logistic.
Another study (Gururaj, H.L., 2023) discussed modifying pixel values and positions based on the
SCAN algorithm and chaotic theory. The SCAN method involved converting image pixel values to new
ones and rearranging pixels in a specific order. Chaotic maps were used to move pixel coordinates within
blocks. Pixel diffusion occurred through the SCAN method, while chaotic maps generated permutations.
Wavelet-based approaches faced limitations due to insufficient phase information, inadequate
directionality, and sensitivity to shifts.
Table 1 provides an overview of the current landscape of various encryption methods utilized in the
field. Referring to specific algorithms and their associated drawbacks, it showcases critical assessments
aiming to enhance the security and functionality of these encryption techniques.
Table 1: An Overview of Various Encryption Techniques
Ref. No
Algorithm
Drawbacks
[19]
AES-ECB
These modifications include reducing computation costs by altering the number of rounds in the MixColumn
transformation, enhancing security by improving the key schedule operation, replacing the S-boxes with a
simpler version to reduce hardware requirements, and addressing pattern appearance problems through ciphering
modes.
[20]
Modified AES
the vulnerability of the modified AES-128 cipher against an impossible differential attack contradicts claims of
improved security. This finding raises concerns about the cipher's reliability and its suitability as a secure
foundation for image encryption, necessitating a reevaluation of its design and potential improvements to ensure
more robust security measures.
[21]
Elliptic Curve
ElGamal
encryption
Inadequate or incomplete standardization of ECEG might lead to interoperability issues or inconsistencies in
different implementations, potentially causing security flaws or vulnerabilities.
[22]
Elliptic curve
cryptography
The paper does not discuss the key generation, distribution, or management processes for ECC-based encryption
in the context of multimedia content.
[23]
Jacobian
elliptic maps
Complexity, implementation, performance, security, algorithmic intricacies, and standardization might pose
issues in their practical usage.
[24]
Elliptic Curve
Diffie-Hellman
key
Weak key management practices could lead to security vulnerabilities
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
168
3 Proposed Model
Recognizing the significance of cryptographic methods in medical imaging, we aim to introduce an
algorithm for a comprehensive encryption strategy. This strategy, known as a Multi-Faceted Encryption
Approach, is designed specifically to secure patient information within medical imaging systems. Its
primary objectives are to adhere to strict data privacy regulations in healthcare, safeguard sensitive
patient data, and uphold the reliability and trustworthiness of these medical imaging systems.
In essence, the proposed algorithm orchestrates a comprehensive encryption process, commencing
with hash code generation to validate image integrity, embedding the hash code through feature-based
watermarking, transforming the image into the frequency domain via DCT, encrypting the frequency
domain image using AES, and culminating in additional security layers through RSA encryption. Each
step contributes significantly to fortifying the image against unauthorized access, tampering, or
interception during transmission, thereby ensuring the confidentiality, integrity, and authenticity of
sensitive medical data contained within the images.
Implementing this algorithm mandates attention to cryptographic best practices, robust
implementation of encryption algorithms, secure key management, and compliance with industry
standards to fortify the confidentiality and integrity of medical images, crucial in safeguarding patient
privacy and maintaining data integrity in healthcare systems.
Medical images hold sensitive and confidential information, necessitating robust encryption
techniques to ensure privacy and integrity during storage or transmission. This algorithm employs a
multi-layered approach, beginning with extracting a cryptographic hash code using the SHA-256
algorithm. The hash code, represented as 'h', is a unique fingerprint for the original image 'X'. This
mathematical representation enables verification of the image's integrity by comparing 'h' before and
after encryption.
Following hash code generation, the algorithm embeds 'h' into the image 'X' through feature-based
watermarking. Feature-based watermarking techniques subtly modify specific image features without
significantly altering their visual appearance. This embedding process, denoted as 'Y =
WatermarkFeature(X, h)', ensures that the hash information becomes imperceptibly integrated into the
image. This step fortifies the image against tampering or unauthorised modifications by concealing the
hash code within its features.
Subsequently, the algorithm transitions the image 'Y' into the frequency domain using the Discrete
Cosine Transform (DCT), denoted as 'Z = DCT(Y)'. This transformation facilitates the conversion of the
image data from the spatial domain to the frequency domain, expressing it as a set of cosine functions.
The use of DCT enables effective representation of image content in a manner conducive to subsequent
encryption processes, laying the groundwork for enhanced security measures.
The encrypted frequency domain image 'Z' undergoes encryption using the Advanced Encryption
Standard (AES) algorithm. AES encryption, employing a symmetric key, 'AES_Key', secures the
transformed image data. The symmetric key is critical in safeguarding the image content, ensuring
confidentiality and preventing unauthorised access to the encrypted data. The resultant encrypted data
'E = AES_Encrypt(Z, AES_Key)' ensures the image's confidentiality and privacy during storage or
transmission.
Further fortifying the encryption, the algorithm applies RSA encryption, a widely used asymmetric
cryptographic technique. RSA encryption involves using public and private keys, ensuring secure
transmission of sensitive information. The encrypted data 'E' is encrypted again using the recipient's
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
169
RSA public key, generating 'F = RSA_Encrypt(E, RSA_Public_Key)'. This multi-layered encryption
enhances the security posture of the medical image by employing both symmetric and asymmetric
encryption mechanisms.
Encryption Algorithm
Step1: Read Medical Image (X)
Step 2: Generate a cryptographic Hash(h)
H=Hash(X)
• 'X' is the original image.
• 'h' represents the hash code generated using a cryptographic hash function (e.g., SHA-256).
Step 3: Apply WaterMarking for embedding Y
Watermarking Algorithm : Y = Watermark(X,h)
• 'Y' results from embedding the hash code 'h' into the image 'X' using watermarking techniques.
• The watermarking algorithm could involve altering pixel values or image features to encode the
hash information.
Step 4: Convert Image to Frequency Domain (Z):
Frequency Transform: Z=FrequencyTransform(Y)
• The operation transforms the image 'Y' into the frequency domain, representing it as 'Z'.
• Mathematical transformations like Discrete Fourier Transform (DFT) or Discrete Cosine
Transform (DCT) might be used.
Step 5: Apply Encryption Algorithm (E):
Encryption: E=Encrypt(Z,K)
• 'E' denotes the result of applying an encryption algorithm to the frequency domain representation
'Z' using a key 'K'.
• This step might involve symmetric encryption (e.g., AES) or homomorphic encryption for
securing the image data.
Step 6:Apply Key-Based RSA Encryption (F):
RSA Encryption: F=RSAEncrypt(E,RSAPublicKey)
• 'F' represents the encrypted data obtained by applying RSA encryption to 'E' using the public
RSA key.
• RSA encryption secures 'E' using the recipient's RSA public key.
Step 7:Output Encrypted Image (F)
Decryption Algorithm
Step 1:Read the Encrypted Image (F):
Step 2: Apply Key-Based RSA Decryption (E'):
RSA Decryption:
E′=RSADecrypt(F,RSA_Private_Key)
• 'E'' represents the decrypted data obtained by applying RSA decryption to 'F' using the recipient's
RSA private key.
• RSA decryption retrieves the AES encrypted frequency domain data 'E'.
Step 3: Apply AES Decryption Algorithm (Z'):
AES Decryption:
Z′=AESDecrypt(E′,AES_Key)
• 'Z'' denotes the decrypted frequency domain image 'Z' using AES decryption with the symmetric
key 'AES_Key'.
• AES decryption retrieves the transformed image data in the frequency domain.
Step 4: Convert Image from Frequency Domain using IDCT (Y'):
IDCT Transformation:
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
170
Y′=IDCT(Z′)
• 'Y'' represents the image 'Y' obtained by inverse Discrete Cosine Transform (IDCT) applied to
'Z''.
• IDCT converts the frequency domain image 'Z'' back to the spatial domain image 'Y'.
Step 5:Verify Watermark and Extract Hash Code (h'):
Watermark Extraction:
ℎ′=h′=ExtractWatermarkFeature(Y′)
• 'h'' denotes the extracted hash code 'h'' from the image 'Y'' using watermark extraction.
• Watermark extraction retrieves the embedded hash information from the spatial domain image
'Y''.
Step 6 Verify Hash Integrity:
Compare Hash Codes:
(ℎ,ℎ′)Match=CompareHashes(h,h′)
• 'Match' indicates whether the original hash code 'h' matches the extracted hash code 'h''.
• Comparison checks the integrity by verifying if the extracted hash 'h'' matches the initially
generated hash 'h'.
Step 7 Output Decrypted Image (Y'):
4 Performance Evaluation
4.1 Analysis
Enforcing this algorithm requires a focus on adhering to best cryptographic practices, ensuring robust
implementation of encryption algorithms, maintaining secure key management, and complying with
industry standards. These efforts are vital in strengthening the confidentiality and authenticity of medical
images, playing a pivotal role in protecting patient privacy and preserving data integrity within
healthcare systems.
Importance of Hashing
The algorithm initiates the generation of a cryptographic hash code using the SHA-256 algorithm. The
significance of hashing lies in its ability to produce a fixed-size string unique to the input data. In the
context of medical images, this hash serves as a digital fingerprint, representing the image without
revealing its contents. This fingerprint, 'h', acts as a reference for image integrity. By comparing the hash
values before and after encryption, any alterations or unauthorized modifications to the image can be
detected.
Integration of Watermarking Techniques
The subsequent step involves embedding the hash code 'h' into the image using feature-based
watermarking. This technique strategically alters specific features or attributes within the image,
effectively concealing the hash information while preserving the image's visual quality. Feature-based
watermarking ensures that the embedded data remains imperceptible to the human eye, safeguarding the
integrity of the image against potential alterations or unauthorised access.
Transformation into the Frequency Domain
Transitioning the image 'Y' into the frequency domain via Discrete Cosine Transform (DCT) provides
several advantages in encryption. DCT converts the image data from the spatial to the frequency domain,
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
171
facilitating better representation and manipulation of the image content. By transforming the image into
a set of cosine functions, it prepares the data for subsequent encryption techniques, enhancing the
image's resilience against potential attacks.
Strengthening Security through AES Encryption
The encrypted frequency domain image 'Z' undergoes Advanced Encryption Standard (AES) encryption,
a symmetric key encryption method known for its robust security measures. AES employs a shared
secret key ('AES_Key') to encrypt and decrypt data, ensuring confidentiality during transmission or
storage. The use of AES enhances the security posture of the image data, preventing unauthorised access
and maintaining confidentiality, which is crucial in protecting sensitive medical information.
Multi-Layered Security via RSA Encryption
To further fortify the encryption, the algorithm employs RSA encryption, an asymmetric cryptographic
technique that adds an extra layer of security. RSA encryption involves a pair of keys (public and
private), with the public key used for encryption and the private key for decryption. By encrypting the
already encrypted data 'E' with the recipient's RSA public key, denoted as 'F = RSA_Encrypt(E,
RSA_Public_Key)', the algorithm enhances data security, ensuring confidentiality even if intercepted
during transmission.
4.2 Simulation
The efficiency of the suggested encryption technique was assessed through practical testing. This
evaluation was conducted on an Intel Core i9 system with 16 GB RAM and a 3.70 GHz processor. A set
of 300 digital medical images, comprising six categories with 50 medical images each (including
Magnetic Resonance Imaging (MRI), Computed Tomography (CT), X-ray scans., and Ultrasound
images sized at 512 × 512), underwent scrutiny. Mecial images were obtained from the "National Library
of Medicine’s Open Access Biomedical Images Search Engine" at https://openi.nlm.nih.gov. Figure 1
showcases a compilation of medical images utilized to analyze and appraise the proposed encryption
method. These images serve as the foundation for assessing how well the proposed encryption system
works and whether it is appropriate for use in the field of medical imaging.
Figure 1: Medical Images Used for the Analysis of the Proposed Scheme
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
172
Figure 2 showcases both the original image (a) and its embedded counterpart (b), visually
representing the watermarking process. The original image (a) serves as the baseline, while the
embedded image (b) demonstrates subtle modifications where the watermark has been seamlessly
incorporated into the visual content. This comparison visually illustrates how the watermarking
technique subtly alters specific image features without significantly impacting their appearance.
Additionally, the corresponding histogram accompanying the images reveals the distribution of pixel
intensities, highlighting any changes introduced by the embedding process. The histogram aids in
quantitatively assessing the alterations in pixel values caused by the watermarking, offering insights into
the impact on the image's tonal distribution and potential perceptual changes. Overall, Figure 2 provides
a comprehensive visual and quantitative depiction of the embedded watermark within the image,
shedding light on the effectiveness of the watermarking process while preserving the image's visual
integrity.
Figure 2: The Original (a) and the Embedded (b) Image and the Corresponding Histogram
Figure 3 compares pixel intensity distributions between the original medical image and the encrypted
version after applying the RSA encryption method. This visualisation serves as a crucial analysis tool,
shedding light on the alterations in pixel intensity values resulting from the encryption process. By
juxtaposing the histograms of the original and encrypted images, this figure offers a clear insight into
how the encryption scheme impacts the pixel intensity distribution. The comparison enables a
quantitative assessment of potential changes in the image's tonal range, highlighting any shifts or
variations introduced by the RSA encryption. Analysing these pixel intensity distributions aids in
evaluating the fidelity of the encrypted image concerning the original, ensuring that while sensitive
medical data is protected through encryption, the essential diagnostic information within the image
remains intact. Figure 3's visual representation of pixel intensity comparisons provides valuable insights
into the encryption's impact on image characteristics, facilitating a comprehensive understanding of the
algorithm's performance in preserving confidentiality and image integrity within healthcare systems.
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
173
Figure 3: The Pixel Intensity Compares the Original and Encrypted Image After Applying the RSA
4.3 Comparison
This section delves into a comprehensive comparison of various metrics utilised to evaluate the
performance and effectiveness of medical image encryption algorithms. In image encryption, assessing
the fidelity, security, and robustness of encrypted images is crucial. To this end, several quantitative
measures and statistical analyses are employed to gauge the quality and safety of the proposed algorithm.
Peak-Signal-to-Noise Ratio (PSNR): This metric quantitatively assesses encrypted images’ quality
by measuring the mean square error between the original, unencrypted image (ground truth) and the
encrypted image.
NPCR (Number of Pixel Change Rate) and UACI (Unified Average Changing Intensity): These
metrics are pivotal in assessing the level of variation and dispersion within encrypted images. NPCR
measures the percentage of pixel changes, while UACI quantifies the average intensity change between
the original and encrypted images, offering valuable insights into the encryption algorithm's behavior.
Entropy: A fundamental measure of randomness and unpredictability in encrypted data, entropy
serves as a crucial indicator of the effectiveness of encryption techniques. High entropy signifies
increased randomness, making decryption without the proper key significantly more challenging.
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
174
It is apparent that a comparison with analogous studies can be conducted, encompassing diverse
methodologies. This comparison involves assessing the image encryption process by referencing similar
works. Subsequently, Table 2 presents the PSNR, NPCR, UACI, and entropy values, focusing on
average metrics derived from implementations in various related works.
Table 2: Performance Comparison with other Related Works
Research
PSNR
NPCR
UACI
Entropy
Proposed Algorithm
27.88
0.997
0.283
7.985
[19]
27.813
0.986
0.279
7.908
[20]
27.601
0.979
0.279
7.624
[21]
27.705
0.986
0.281
7.654
[22]
27.516
0.980
0.273
7.638
[23]
27.874
0.981
0.280
7.601
[24]
27.850
0.978
0.276
7.781
[25]
27.771
0.983
0.274
7.857
[26]
27.760
0.984
0.278
7.642
5 Conclusions and Future Work
The proposed encryption algorithm has been meticulously developed to fortify the security of patient
information within medical imaging systems. By embracing cryptographic methodologies tailored
specifically for medical imaging, this approach upholds paramount objectives: compliance with stringent
healthcare data privacy regulations, protection of sensitive patient data, and the unwavering assurance
of reliability and trustworthiness in medical imaging systems.
Upon comparison with cutting-edge encryption methods, the suggested encryption approach displays
characteristics akin to a robust cipher. Through a commendable encryption assessment, yielding high
values in Peak Signal-to-Noise Ratio (PSNR), Non-Probabilistic Convergence Ratio (NPCR), and
Unified Average Changing Intensity (UACI), coupled with a substantial entropy value, the proposed
encryption method showcases considerable resilience and effectiveness in preserving the confidentiality
and integrity of critical medical data. This amalgamation of statistical and security analyses reaffirms
the strength and efficiency of the proposed encryption approach, highlighting its potential as a
fundamental component in securing patient information within medical imaging systems.
Author Contributions: Idea development, approach, detailed analysis, research, and initial drafting
by A.O.; A.A. and A.O. contributed to editing and revising the written content. All authors have
reviewed and approved the manuscript for publication.
Acknowledgments: The authors express their genuine appreciation to Princess Sumaya University for
Technology for their assistance and support throughout this project.
Conflicts of Interest: The authors declare no conflicts of interest.
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
175
References
[1] Aram, S., Shirvani, R. A., Pasero, E., & Chouikha, M. F. (2016). Implantable Medical Devices;
Networking Security Survey. Journal of Internet Services and Information Security (JISIS),
6(3), 40-60.
[2] Bandari, V. (2023). Enterprise Data Security Measures: A Comparative Review of Effectiveness
and Risks Across Different Industries and Organization Types. International Journal of
Business Intelligence and Big Data Analytics, 6(1), 1-11.
[3] Behnia, S., Akhavan, A., Akhshani, A., & Samsudin, A. (2013). Image encryption based on the
Jacobian elliptic maps. Journal of Systems and Software, 86(9), 2429-2438.
[4] Cheng, Z., Wang, W., Dai, Y., & Li, L. (2022). A High-Security Privacy Image Encryption
Algorithm Based on Chaos and Double Encryption Strategy. Journal of Applied
Mathematics, 2022.
[5] Dzwonkowski, M., Papaj, M., & Rykaczewski, R. (2015). A new quaternion-based encryption
method for DICOM images. IEEE Transactions on Image Processing, 24(11), 4614-4622.
[6] El-Latif, A.A.A., Ramadoss, J., Abd-El-Atty, B., Khalifa, H.S., & Nazarimehr, F. (2022). A
novel chaos-based cryptography algorithm and its performance analysis. Mathematics, 10(14),
1-22.
[7] Ferdush, J., Begum, M., & Uddin, M.S. (2021). Chaotic lightweight cryptosystem for image
encryption. Advances in Multimedia, 2021, 1-16.
[8] Furukawa, M.F., King, J., Patel, V., Hsiao, C.J., Adler-Milstein, J., & Jha, A.K. (2014). Despite
substantial progress in EHR adoption, health information exchange and patient engagement
remain low in office settings. Health Affairs, 33(9), 1672-1679.
[9] Gold, M., & McLaughlin, C. (2016). Assessing HITECH implementation and lessons: 5 years
later. The Milbank Quarterly, 94(3), 654-687.
[10] Gururaj, H. L., Almeshari, M., Alzamil, Y., Ravi, V., & Sudeesh, K. V. (2023). Efficient SCAN
and Chaotic Map Encryption System for Securing E-Healthcare Images. Information, 14(1), 1-
15.
[11] Hamza, N.A., Jafeer, S.H., & Ali, A.E. (2019). Encrypt 3d model using transposition,
substitution, folding, and shifting (tsfs). In IEEE 2nd Scientific Conference of Computer
Sciences (SCCS), 126-131.
[12] Jouini, M., & Rabai, L.B.A. (2019). A security framework for secure cloud computing
environments. In Cloud security: Concepts, methodologies, tools, and applications, 249-263.
IGI Global.
[13] Kaissis, G.A., Makowski, M.R., Rückert, D., & Braren, R.F. (2020). Secure, privacy-preserving
and federated machine learning in medical imaging. Nature Machine Intelligence, 2(6), 305-
311.
[14] Kumar, M., Iqbal, A., & Kumar, P. (2016). A new RGB image encryption algorithm based on
DNA encoding and elliptic curve Diffie–Hellman cryptography. Signal Processing, 125, 187-
202.
[15] Li, L., Abd El-Latif, A.A., & Niu, X. (2012). Elliptic curve ElGamal based homomorphic image
encryption scheme for sharing secret images. Signal Processing, 92(4), 1069-1078.
[16] Lite, S., Gordon, W.J., & Stern, A.D. (2020). Association of the meaningful use electronic health
record incentive program with health information technology venture capital funding. JAMA
network open, 3(3), 1-10.
[17] Nosrati, S., Sabzali, M., Heidari, A., Sarfi, T., & Sabbar, S. (2020). Chatbots, counselling, and
discontents of the digital life. Journal of Cyberspace Studies, 4(2), 153-172.
[18] Panayides, A.S., Amini, A., Filipovic, N.D., Sharma, A., Tsaftaris, S.A., Young, A., & Pattichis,
C.S. (2020). AI in medical imaging informatics: current challenges and future directions. IEEE
journal of biomedical and health informatics, 24(7), 1837-1857.
A Multi-Faceted Encryption Strategy for Securing Patient
Information in Medical Imaging
Ammar Odeh et al.
176
[19] Pollard, J.M. (1978). Monte Carlo methods for index computation (). Mathematics of
computation, 32(143), 918-924.
[20] Ramzan, M., Habib, M., & Khan, S.A. (2022). Secure and efficient privacy protection system
for medical records. Sustainable Computing: Informatics and Systems, 35.
[21] Rasool, R.U., Ahmad, H.F., Rafique, W., Qayyum, A., & Qadir, J. (2022). Security and privacy
of internet of medical things: A contemporary review in the age of surveillance, botnets, and
adversarial ML. Journal of Network and Computer Applications, 201.
[22] Reisman, M. (2017). EHRs: the challenge of making electronic data usable and
interoperable. Pharmacy and Therapeutics, 42(9), 572-575.
[23] Shukla, S., George, J. P., Tiwari, K., & Kureethara, J.V. (2022). Data Ethics and Challenges.
Springer, 41-59.
[24] Singh, S., Hosen, A.S., & Yoon, B. (2021). Blockchain security attacks, challenges, and
solutions for the future distributed iot network. IEEE Access, 9, 13938-13959.
[25] Tawalbeh, L.A., Mowafi, M., & Aljoby, W. (2013). Use of elliptic curve cryptography for
multimedia encryption. IET Information Security, 7(2), 67-74.
[26] Teh, J.S., Alawida, M., & Sii, Y.C. (2020). Implementation and practical problems of chaos-
based cryptography revisited. Journal of Information Security and Applications, 50, 1-41.
[27] Wadi, S.M., & Zainal, N. (2014). High definition image encryption algorithm based on AES
modification. Wireless personal communications, 79, 811-829.
[28] Yao, F. (2021). Hybrid Encryption Scheme for Hospital Financial Data Based on Noekeon
Algorithm. Security and Communication Networks, 2021, 1-10.
[29] Yap, W.S., Phan, R.C.W., & Goi, B.M. (2016). Cryptanalysis of a high-definition image
encryption based on AES modification. Wireless Personal Communications, 88(3), 685-699.
[30] Yu, F., Qian, S., Chen, X., Huang, Y., Cai, S., Jin, J., & Du, S. (2021). Chaos-based engineering
applications with a 6D memristive multistable hyperchaotic system and a 2D SF-SIMM
hyperchaotic map. Complexity, 2021, 1-21.
[31] Zhang, J., Li, G., Marshall, A., Hu, A., & Hanzo, L. (2020). A new frontier for IoT security
emerging from three decades of key generation relying on wireless channels. IEEE Access, 8,
138406-138446.
Authors Biography
Ammar Odeh received his Ph.D. Degree in Computer Science and Engineering with a
concentration in Computer Security (Steganography) from the University of Bridgeport. He
received an M.S. in Computer Science with a concentration in Reverse Software Engineering
and Computer Security from the University of Jordan, College of King Abdullah II School for
Information Technology (KASIT). In 2002, he finished his B.Sc. Degree in Computer Science
and applications from the Hashemite University, Prince Al-Hussein Bin Abdullah II for
Information Technology. During his Ph.D., he worked as a Research Assistant, Teaching
Assistant, and Instructor. He is currently an assistant professor in computer science at Princess
Sumaya University for Technology.
Anas Abu Taleb is an associate professor in the Department of Computer Science at Princess
Sumaya University for Technology, Amman, Jordan. He received a Ph.D. in Computer Science
from the University of Bristol, UK, in 2010, an MS.c. in Computer Science from the University
of the West of England, UK, 2007 and a BS.c. Degree in Computer Science from Princess
Sumaya University for Technology, Jordan, 2004. Dr. Abu Taleb has published several journal
and conference papers on sensor networks. In addition to sensor networks, Dr. Abu Taleb is
interested in network fault tolerance, routing algorithms, and mobility models.