Privacy-preserving self-serviced medical diagnosis scheme based on secure multi-party computation

With the development of the "Internet + Intelligent Medical", patients can online diagnose some common diseases via the Internet. However, during the diagnostic process, there exist many severe problems on privacy for medical sensitive data of patients. To solve these problems, in this paper, we present a new privacy-preserving self-serviced medical diagnosis scheme based on secure multi-party computation (SMC). In our scheme, a registered patient first encrypts his/her medical health data and sends it to the hospital server, then the hospital server calculates the similarity value between the patient's medical health data and the trait vector of hospital disease. Finally, the hospital server searches for the disease that matches the patient according to the calculated similarity value, and sends the treatment method of this disease to the patient. Specifically, based on homomorphic encryption (HE) and privacy-preserving access control, our self-serviced medical diagnosis scheme can achieve privacy preservation of patient's medical health data and confidentiality of hospital diagnosis mode. Through detailed security analysis, we show that our scheme can resist various known security threats. In addition, our scheme not only reduces the cost of treatment for the patients and relieves the hospitals' heavy pressure in the course of diagnosis, but can also predict other diseases of the patients, which can make the patients a more clear understanding of their current physical condition, and the patients can obtain the most appropriate treatment. (C) 2019 Elsevier Ltd. All rights reserved.