Smart Healthcare Data Privacy Enhancement Using PSK-ECC, STh-RNN, Pseudonymization, and Data Diddling
Sandeep Pandey, P. Elamurugan, M. Shakunthala, Amit Lathigara, Kakarla Ramana Reddy, N. Thulasi Chitra · 2025
A further variation of the hashed methodology, a data masking method, and methods to further enhance data security during sensing and storage while reducing computational expenses caused by encryption are introduced. Data identity theft, role-based consent, and information v-diddling identity are all presented in this study with the goal of enhancing the security of data. Initially, those who utilize IoT data had to register with the hospital's cloud-based server. Then, in order to avoid assaults such as the birthday attack, the 2CBA (2's Addition Binary Additions) approach was used to transform the users' position and IP address into a number that is binary. After that, the user's login information is entered into their user profile database so that the STh-RNN algorithms may be trained to detect data manipulation. Once the data user confirms the request, the authorization server will produce the code by using the HRAS-BH approach, which stands for Hashed Role-based Accessing Structural utilizing Blake Hashing. In comparison with additional methods such as Convolutional Neural Network (CNN), Deep Neural Network (DNN), Deep Belief Network (DBN), and Recurrent Neural Network (RNN), the suggested Public Secret Key - Elliptic Curve Cryptography with Swish Tanh - Recurrent Neural Network (PSK-ECC with STh-RNN) solution gets the best accuracy at 98.13%. This proves that the suggested technique is a more trustworthy strategy for safeguarding and handling medical information, as it far surpasses the current techniques in precisely this regard.