Comparative Study on Encryption Algorithms for Autism MRI Scan Images

Abstract

Healthcare centre is the most susceptible sctor. Cyber-attack on medical institutions, such as hospitals and health care centre have increased year after year. Autism health care centre is one of the victims. Autism MRI scan images include sensitive and confidential information about a patient's health and other personal information. Therefore, image encryption is a fundamental aspect of information security, especially in today's digital era where the proliferation of images and the need to protect sensitive visual content are of paramount importance. In order to protect the confidentiality and integrity of images, image encryption techniques work to make sure that only authorised users can view and understand the visual data while prohibiting unauthorised access and modification. The pixel values and spatial properties of the image are hidden using a variety of cryptographic methods and key-based actions. Symmetric encryption algorithm, like Advanced Encryption Standard (AES), which use the same key for encryption and decryption operations, are among the famous image encryption techniques. After that, asymmetric encryption algorithm like Rivest-Shamir-Adleman (RSA), which employ two distinct keys for both encryption and decryption. Next, the research includes a hybridization method of Elliptic Curve Cryptography (ECC) with the Hill Cypher (HC) for image encryption and decryption. There are five different Autism MRI Scan images selected to be used as the input image for encryption process. Then, each image will have two different file format which is PNG and BMP and two different file size which is 256*256px and 512*512px. This research's goal is to compare how well symmetric encryption, asymmetric encryption and hybrid encryption work as three separate image encryption techniques. This can help users rapidly and efficiently speed up the encryption and decryption of images and enhance the security of MRI scan images. By applying differential analysis parameters like Number of Pixel Change Rate (NPCR), Unified Average Changing Intensity (UACI) and statistical analysis parameters like Histogram Analysis (HA), the characteristics of the image encryption approach are examined. At the end of the experiment, AES encryption proves to be the best among the three types of encryption with better performance in terms of security with NPCR percentage of 99.72%, UACI percentage of 32.68% and has a higher speed of encryption which only takes 0.0238s but has the second-best distribution in HA of pixel value to frequency compares to ECC with HC.