Towards Optimizing the Usability of Homomorphic Encryption in Cloud-Based Medical Image Processing

Undoubtedly, digital imaging continues to play a crucial role in the diagnosis and treatment of disease. In recent years, there has been growing interest in using Cloud services in the healthcare domain. Even though Cloud Computing offers many advantages, the shift to this new paradigm still poses many problems relating to privacy and data protection. Therefore, it is significantly important to revisit the issue of medical image processing adopting a new method namely Cloud services. We place heavy emphasis on homomorphic encryption schemes, which play a useful role in dealing with privacy and security challenges. In reality, running time is one of the major disadvantages of these types of algorithms. In this context, the present contribution can be conceived of as a novel approach to improve both performance and security. This is achieved by using Multi-Agent System (MAS) and encrypting only the content of a Region-of-Interest (RoI). This would dramatically minimize the size of the needed data. Consequently, it improves the response time of the proposed framework. Furthermore, the generated RoI is divided into certain number of shares, and hence, processed by different agents implemented in several nodes. Following this, this approach is meant to enhance both confidentiality and performance. The experimental results prove that this method is an efficient solution for data encryption in encrypted domain. In fact, the proposal is meant to reduce the running time required to process images that are encrypted using the homomorphic algorithms.