In the healthcare system, the application of predictive analysis is essential to the enhancement of patient benefits as well as the development of healthcare delivery systems. The digitization of health records presents an increasing threat of data leakage and breaches of patient privacy. This chapter discusses how homomorphic encryption can be applied as a solution to secure healthcare information. We discuss the state of the art of predictive analysis in healthcare organizations and realize that the issue of data security is still relevant and needs to be further investigated due to evolving healthcare regulations and rapid technological advancements. We then discuss an overview of different available encryption techniques. We particularly focus on homomorphic encryption that allows computations to be made on data without decryption while maintaining patient data privacy. After that, we discuss how predictive analysis techniques can be applied to encrypted healthcare data. Some of the issues arising when attempting to carry out predictive analysis on encrypted data are discussed, in addition to the advantages of and technical hurdles in homomorphic encryption. We examine trends and opportunities, focussing on how secure predictive analytics, as one of the potential solutions, can improve the trust and reliability of healthcare data and patients’ care. Finally, we perform a case study on the use of predictive analysis techniques in encrypted heart disease data with the help of the Paillier Homomorphic encryption scheme to maintain data security.
Keywords: Case studies, Data security, Homomorphic encryption, Machine learning model, Privacy, Secure healthcare predictive analysis.
