Enhanced Medical Image Steganography Using Improved LSB With Conditional MSB Based on Color Vector Variety

Abstract

Image steganography involves concealing various forms of information, such as text, images, or videos, within an existing image. The hidden data is embedded in a manner that remains imperceptible to human observers. In this paper, a stego image that successfully hides the secret information and renders it undetectable to others is created by embedding a hidden message into an original image. The secret message may be text (the doctor’s report of the patient’s case) or maybe a secret image (chest x-ray) or encrypted secret image to increase security and guarantee the patient’s privacy. The cover images may be brain, breast, and lung images. An enhanced LSB technique is introduced, incorporating common LSB and conditional MSB, with a focus on the lowest energy (blue channel) in the RGB spectrum. This approach aims to maintain image quality and imperceptibility. The innovation lies in strategically combining conditional MSB with the selective use of the blue channel, enhancing security. The method demonstrates superior concealment of secret information, surpassing conventional LSB techniques. Additionally, the Structural Similarity Index (SSIM) reports a high value of 0.998, affirming the technique's success in concealing information while preserving image quality. Performance metrics show positive results due to minimal changes (2 LSB bits at most). The experiments demonstrated positive outcomes in comparison to prior studies in this field.