Enhanced Privacy Preservation Technique for the Multi Institutional Clinical Data Using Hybrid Feline-Storm Algorithm

Sagarkumar Patel, Rachna Patel

Today's medical research is seen to be highly dependent on data exchange; unfortunately, despite its benefits, it frequently encounters problems, particularly issues with data privacy. As a result, several methods and infrastructures have been created to ensure that patients and research participants maintain their anonymity when data is exchanged. However, privacy protection often has a cost, such as limitations on the types of studies that may be done on shared data. The lack of a systematization that would make the trade-offs made by various techniques obvious is what needs to be addressed. In this research, develop the Feline-Storm Based Privacy Preservation Technique for multi-institutional clinical data. Data mining provides many advantages in various domains, particularly in medicine. The data about the disease is ensured to the experts, who can determine the effects, availability, and nature. The private information of the persons should not be disclosed to the expert groups, which ensures the confidentiality of the confidential information. Hence, to ensure the privacy of the people's electronic health records (EHR), this research utilizes the C-mixture and three privacy restraints that strengthen the privacy measures. Furthermore, the Hybrid Feline-storm algorithm, which emphasizes exploitation or the exploration phase at any instance, avoiding the local optima and the premature convergence to ensure the optimized privacy preserved of the data. This research also establishes security strategies such as K-anonymity, T-closeness, and L-diversity to attain complete data privacy. Further, the Feline-storm optimization is developed to minimize information loss. The information loss, class average size, and fitness measure achieved by the proposed methodology are 0.85, 0.38, and 4.7457, respectively.