A Novel Medical Image Encryption using Cyclic Coding in Covid-19 Pandemic Situation

During this Covid - 19 pandemic situation, encryption of medical images takes a major role in medical information systems as well as in telemedicine. However according to government rule, it is essential to hide the information of the patients. Recent development in computer network enhances a lot of facilities in communication area. Unfortunately, hackers are misusing this facility and always try to attack on the transmitted information in insecure network. For secure transmission of medical images, it is essential to encrypt the information before transmitting. In this communication, we are going to present a novel approach of medical image encryption using cyclic coding. We have proved that it is quite difficult to decrypt the original information from encoded data in one common mode of attacking-chosen ciphertext attack. Moreover, we have proved the effectiveness of the encryption using correlation coefficients. Our proposed scheme is suitable for efficient encoding of multiple medical images.


Introduction:
In general, images used in medical field are treated as sensitive information in the bio information systems. To transmit these types of images through network, a secure encryption algorithm is required.
Researchers have proposed a lot of image encryption algorithm for the last few decades. Data Encryption Standard algorithm is oldest algorithm used for encryption which requires minimum time for cost computations [1,2]. By applying either symmetric or asymmetric algorithm, a secure image encryption can be done in very effective way. [3][4][5]. Image encryption is considered very effective while it can provide better result against common attacks-known-plaintext as well as ciphertext attack [6]. In medical image encryption, to enhance security, Cipher Feedback Mode is used to encrypt the image [7].
Shamir described a secret sharing encryption scheme based on (k, n) method and polynomial  [9]. In this method, with the proper value of k and n, the selected images can be reconstructed using the retrieval process.
In 2008, Hill explained an encryption method at both grayscale and color images [11] with a drawback of same level of color attributes on background. Nag described a method of rearranging the pixel position within the specified target images with digital function, XORing [12]. This method shows an excellent correlation information. Sokouti and his associates [13,14], proposed a one time password based encryption system by splitting the image bits into block of rows for encryption purpose. With the help of pixels block and permutation, encryption has been performed by Rijn Dael algorithm process [15]. Ismail described [17] an algorithm based large key space to make more different pixels. Kamali SH [18] has been proposed modified version of AES where the security are highly increased in comparison to normal system of AES. A Patra, A Saha and K Bhattacharya proposed an effective image encryption algorithm using phase grating along with random phase matrix. They have analysed their values with plaintext and ciphertext attacks. [19]. A lot of different methods are successfully used in to hide images, in both spatial and frequency domain [20][21][22]. LSB bit replacement technique, a wellknown method, widely used for its embedding property which is really useful for changing information of a bit in each pixel [23]. A Patra proposed a method of using sinusoidal phase grating and random phase masking for multiplexing, compression and encryption of digital images [24][25][26][27] In this present research work, medical images are encrypted using well known cyclic coding and convolution coding. In general, there two types of cyclic coding-non-systematic and systematic cyclic coding. We have used non-systematic cyclic coding. A well-known generator polynomial is selected encoding as well as decryption process. Initially the selected images are multiplied by generator polynomial for encryption and at the receiving end they are decrypted using inverse process. Our work is mainly in frequency domain; therefore, some spatial-domain cyber-attacks are not effective. In our proposed method, decryption of a particular image is dependent on proper generator polynomial selection. To check the security of tour method, we have applied this in very well-known cyber-attack technique-chosen ciphertext attack. Correlation coefficient is one effective tool to measure the effectiveness in image encoding process. In every case, correlation coefficient provides satisfactory results.

Methodology:
The selected image is represented by f (x, y). We have used non-systematic cyclic coding. In nonsystematic cyclic coding system, generator polynomial The entire encoded image is transformed into frequency domain.
It should be mentioned that Mod 2 arithmetic is applied here. .

Result:
In this section we have showed our results. We have worked with MATLAB software. Correlation coefficient of the output images is also shown in the table.
Selected images are displayed in Fig 1 (a-c). Pixel resolution of each medical image is 512 x 512.

Chosen Ciphertext Attack :
We have decrypted the encrypted images using different keys. In each case, correlation coefficients have been calculated of the original and decrypted plaintexts for all images. Though we have applied this process on every image using different keys, only three cipher texts have been displayed in the Fig. 4 (a-c) and correlation coefficient test is mentioned in Table 2.   In absence of anyone parameter, it is quite impossible to decode the original image. The proposed method is very fast and it is suitable for encryption of multiple medical images in telemedicine area. In preventing piracy of digital images, this method also can be effectively used.