A Privacy-Preserving Online Medical Prediagnosis Scheme for Cloud Environment

Abstract

The paradigm of online medical prediagnosis has emerged to ease the shortage of health professionals in rural areas. It can provide a 24-hour online health care service and guide rural residents' medical treatment. However, the development of online medical prediagnosis system still faces many challenges, involving the leakage and overuse of medical information. In this paper, we utilize the logistic regression to design a privacy-preserving medical prediagnosis scheme for the cloud environment, named POMP, which provides a health care service for users without violating their privacy. It is characterized by employing homomorphic encryption techniques to achieve a privacy-preserving prediagnosis process over the encrypted data. The proposed POMP scheme also adopts a preprocessing technique and Bloom filter to reduce the computational cost in the prediagnosing process. Through extensive analyses, we demonstrate that the proposed POMP scheme can resist various security threats and protect the privacy successfully. In order to evaluate the performance, we also implemented the POMP scheme and measured the running time on the smartphone and computer. The experimental result shows POMP's efficiency in terms of the computational and communication burden.

Keywords

• Homomorphic encryption
• logistic regression
• online medical prediagnosis
• privacy preservation