Research on Information Security of Pulmonary Nodules Based on DOSA

Abstract

At present, the information disclosure, especially medical information disclosure, is very serious, According to bitglass, the number of medical data leaks in the United States doubled in 2020, affecting more than 26 million people.this paper proposes an information security model based on DOSA (data-oriented security architecture [1]). this module integrates various existing pulmonary nodule algorithms, the data is triple protected based on the DOSA[2]. Firstly, the elliptic curve digital signature algorithm is used to encrypt the data, then the Tang poetry dictionary is divided into a hash table by Tang poetry encryption algorithm, and the hash table is converted into bits according to the rhythm of Tang poetry to generate encrypted text; If the first and second parts are cracked and destroyed by malicious users, the system will utilize convolutional neural network and a deep learning algorithm to perform convolutional confrontation algorithm. for data security and efficiency, the severs will utilize the load balancing to process the input data.experiments show that, the information security model in this paper can keep the protection rate against malicious attacks above 90% without affecting the server performance.