Quality optimized medical image information hiding algorithm that employs edge detection and data coding
Related Experiment Video
Updated: Mar 23, 2026
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Hayat Al-Dmour1, Ahmed Al-Ani1
1School of Electrical, Mechanical and Mechatronic Systems, Faculty of Engineering and Information Technology, University of Technology Sydney, NSW 2007, Australia.
Objectives:
The present work has the goal of developing a secure medical imaging information system based on a combined steganography and cryptography technique. It attempts to securely embed patient's confidential information into his/her medical images.
Methods:
The proposed information security scheme conceals coded Electronic Patient Records (EPRs) into medical images in order to protect the EPRs' confidentiality without affecting the image quality and particularly the Region of Interest (ROI), which is essential for diagnosis. The secret EPR data is converted into ciphertext using private symmetric encryption method. Since the Human Visual System (HVS) is less sensitive to alterations in sharp regions compared to uniform regions, a simple edge detection method has been introduced to identify and embed in edge pixels, which will lead to an improved stego image quality. In order to increase the embedding capacity, the algorithm embeds variable number of bits (up to 3) in edge pixels based on the strength of edges. Moreover, to increase the efficiency, two message coding mechanisms have been utilized to enhance the ±1 steganography. The first one, which is based on Hamming code, is simple and fast, while the other which is known as the Syndrome Trellis Code (STC), is more sophisticated as it attempts to find a stego image that is close to the cover image through minimizing the embedding impact. The proposed steganography algorithm embeds the secret data bits into the Region of Non Interest (RONI), where due to its importance; the ROI is preserved from modifications.
Results:
The experimental results demonstrate that the proposed method can embed large amount of secret data without leaving a noticeable distortion in the output image. The effectiveness of the proposed algorithm is also proven using one of the efficient steganalysis techniques.
Conclusion:
The proposed medical imaging information system proved to be capable of concealing EPR data and producing imperceptible stego images with minimal embedding distortions compared to other existing methods. In order to refrain from introducing any modifications to the ROI, the proposed system only utilizes the Region of Non Interest (RONI) in embedding the EPR data.
