Only array. Rows of 2Darray were assigned odd
Only a few research papers for medical image securitybased on watermarking and encryption approaches exist.Below we discuss in brief a few recently developed medicalimage watermarking and encryption schemes. Coatrieuxet al. (2007) proposed a reversible and regions of noninterest(RONI) watermarking for the medical image reliabilityprotection. Velumani and Seenivasagam (2010) proposeda reversible blind watermarking scheme to watermarka medical image with the patient’s photograph in an invisiblemanner such that it is available to the respective physicianonly on extraction with a key. Lin et al. (2010) proposed amethod to authenticate and protect medical images in theregion of interest (ROI) whereas rest of the ROI was used forembedding watermark that contained the patients’ data andauthentication information. The authentication informationwas generated from the ROI by analyzing wavelet coefficientswith singular value decomposition. Lin et al. (2012)proposed amethod to hide the patient’s information intomedicalimages based on singular vale decomposition (SVD). Inthis method, SVD was employed to find features of medicalimage and then the information of patient was hiddenin the non-ROI area of the image. Goyat (2012a) exploreddifferent techniques of cryptography to prove that the naturalselection-based techniques are as good as the rigorousmathematical techniques. Further, Goyat (2012b) suggesteda new approach to generate secret key for encryptionalgorithm based on GAs. Khan and Bhatia (2012) suggestedan algorithm to generate key using Genetic Algorithms.In the algorithm, a population was generated usingpseudorandom number generator. Each chromosome wasconsidered the size of 25 cells and 1,000 chromosomeswere taken in their experiment. They divided each chromosomeinto five groups and calculate number of ones ineach group. Depending on number of ones, 25 × 1,000 chromosomeswere transformed into 5 × 1,000 chromosomesand again further, they were converted into 25 × 200 chromosomes.Further, crossover and mutation operators wereapplied on final population. Dutt et al. (2013) recommendedimplementing security using Genetic Algorithms. In theirsuggested text encryption algorithm, a fitness function wasdefined which was used to assign weightage to each alphabetuniquely depends on the position of alphabet. Further, ablock of characters were read into 2D array. Rows of 2Darray were assigned odd random number between 1 and9 and columns were assigned fitness function weightage.Each matrix elements were sum up of corresponding rowand column values. Further, from 2D array, maximum valuewas selected and omits corresponding row and column elements.The process was repeated several times in their algorithm.Fu et al. (2013) proposed a chaos-based medical imageencryption scheme. The proposed scheme introduced a substitutionmechanism in the permutation process through abit-level shuffling algorithm. Afarin and Mozaffari (2013)developed an image encryption using Genetic Algorithm. Intheir approach, initially rows and columns of plain imageare reshuffled and then divided resultant image into fourequal-sized sub-images. Take sub-image one by one and GAoperations are applied on them. Entropy was chosen as fitnessfunction in their approach. Pizzolante et al. (2013) proposeda hybrid approach that handles simultaneously andefficiently both the compression that the security of threedimensionalimages wherein a low complexity approach forthe compression of 3-D medical images followed by an efficientmethod to insert, within each image, an invisible digitalwatermark during the compression process is used. Castiglioneet al. (2014) proposed an engine for lossless dynamicand adaptive compression of 3D medical images, which alsoallows the embedding of security watermarks within them.Furthermore, to provide effective, secure and flexible accessto healthcare resources that need to be managed by medicalapplications, an architecture of a SaaS Cloud system has beendefined. This allows devices with totally different and heterogeneoushardware and software characteristics to interactamong them making them completely transparent to the enduser.