Pixel-based dispersal scheme for medical image survivability and confidentiality

Abstract

Digital medical image survivability and confidentiality is an essential concern for healthcare institution in archiving medical image. Survivability issue in the current conventional archiving system happened when the server becomes temporarily inaccessible; meanwhile confidentiality happened when the data disclosed by unauthorized party. This project introduces a potential approach by integrating Rabin’s Information Dispersal Algorithm (IDA) and Shamir’s Secret Sharing Algorithm (SSA) for solving these issues. These approaches have been already known for providing data survivability and confidentiality in other fields. This project applies these secret sharing threshold schemes to a system called Medical Image Dispersal Picture Archiving Communication System (MIaDPACS) to provide digital medical image survivability and confidentiality. The pixels are extracted, encoded, and dispersed by using Rabin’s IDA into distributed archiving system. The secret key used in Rabin’s IDA is dispersed by using Shamir’s SSA. Original image reconstruction requires only subset of the dispersed files with numbers equals to a defined threshold during dispersal. Experimental results showed that the reconstructed image was exactly identic with the original image both for nosalt and salt scheme. This is shown by Peak Signal to Noise Ratio (PSNR) value that is equal to infinity. In addition, random dispersed files does not reveal the original image. Thus, MIaDPACS is able to provide the survivability and confidentiality for digital medical image.