Secure Medical Data Transmission: An Adversarial Neural Cryptography-Based Steganography Technique with Digital Signature and LSB Replacement

Abstract

It is difficult to secure sensitive healthcare data in light of current technological advancements.

Protecting medical material, including patient privacy, is one of the most important concerns with

medical information security. In order to safeguard confidentiality, integrity, and availability,

security measures must be put in place as medical information becomes increasingly accessible.

Recently, image steganography was suggested as an additional data security measure for medical

information. This study describes a data-hiding strategy for DICOM medical images. In order to

maintain confidentiality, we encrypt and hide the RGB patient image within the Region of Non

Interest (RONI) of the medical image using Adversarial Neural Cryptography with SHA-256

(ANC-SHA-256). Before embedding, we encrypt the RGB patient picture using ANC-SHA-256

to guarantee anonymity. To verify the integrity and authenticity of medical images, we use a secure

hash technique with 256 bits (SHA-256) to create a digital signature from the data associated with

the DICOM file. Using a variety of medical datasets, such as MRI, CT, X-ray, and ultrasound

cover images, several tests were carried out to evaluate visual quality. As a patient-concealed

image, the LFW dataset was used. In terms of visual quality metrics, such as the PSNR average of

67.55, the NCC average of 0.9959, the SSIM average of 0.9887, the UQI average of 0.9859, and

the APE average of 3.83, the suggested approach works well. In six medical evaluation categories,

it performs better than the state-of-the-art methods in these visual quality metrics (PSNR, MSE,

and SSIM). Additionally, the suggested approach provides excellent visual quality while

withstanding geometrical threats like cropping, rotation, and scaling, as well as physical changes

and histogram analysis. Lastly, its high attaining security ratio of 99% during remote transmission

of Electronic Patient Records (EPR) via the Internet makes it especially effective in telemedicine

applications, protecting patient privacy and data integrity.