Steganography Algorithms in Computed Tomography (CT) Scan Images

Abstract

Steganography is a technique which allows to hide information or data inside a medium without giving an idea to people that the hidden data even existed. In the medical field, there are cases where patient’s medical images breached along with their sensitive and personal data because of poor security. This problem leads to a wanting of a good solution to have the medical image with secured patient data. Here is where the image steganography takes place to hide the patient data inside the medical images. This study is conducted to determine the effective image steganography algorithm to be used in the medical field to embed the patient information inside their medical images specifically in Computed Tomography (CT) scan images. The image steganography algorithms that will be used in this study are Least Significant Bit (LSB), Exploiting Modification Direction (EMD) and Discrete Wavelet Transform (DWT). The flow of the image steganography technique will be studied and analyzed and after that, the algorithms will be implemented with different file formats which are Portable Network Graphics (PNG), Bitmap (BMP) and Joint Photographic Experts Group (JPG) and different image sizes which are 1024x1024 and 2048x2048 pixels. After that, the results are evaluated with performance metrics which are Peak Signal-to-Noise Ratio (PSNR), Mean Squared Error (MSE), Structural Similarity Index Measure (SSIM) and Pairs Analysis. The results are then compared to determine the best image steganography algorithm along with the image file format which gives the highest robustness, visual quality, security and imperceptibility. The research result displays the EMD algorithm as the best paired with BMP file format in image steganography where it resulted in producing highest average of SSIM and PSNR which is 1.0 and 67.421 respectively and lowest MSE average of 0.012 with no traces in pairs analysis.