Privacy-Preserving Patient-Centric Clinical Decision Support System on Naïve Bayesian Classification

Clinical decision support system, which uses advanced data mining techniques to help clinician make proper decisions, has received considerable attention recently. The advantages of clinical decision support system include not only improving diagnosis accuracy but also reducing diagnosis time. Specifically, with large amounts of clinical data generated everyday, naïve Bayesian classification can be utilized to excavate valuable information to improve a clinical decision support system. Although the clinical decision support system is quite promising, the flourish of the system still faces many challenges including information security and privacy concerns. In this paper, we propose a new privacy-preserving patient-centric clinical decision support system, which helps clinician complementary to diagnose the risk of patients' disease in a privacy-preserving way. In the proposed system, the past patients' historical data are stored in cloud and can be used to train the naïve Bayesian classifier without leaking any individual patient medical data, and then the trained classifier can be applied to compute the disease risk for new coming patients and also allow these patients to retrieve the top-k disease names according to their own preferences. Specifically, to protect the privacy of past patients' historical data, a new cryptographic tool called additive homomorphic proxy aggregation scheme is designed. Moreover, to leverage the leakage of naïve Bayesian classifier, we introduce a privacy-preserving top-k disease names retrieval protocol in our system. Detailed privacy analysis ensures that patient's information is private and will not be leaked out during the disease diagnosis phase. In addition, performance evaluation via extensive simulations also demonstrates that our system can efficiently calculate patient's disease risk with high accuracy in a privacy-preserving way.

I. Introduction

Healthcare industry, extensively distributed in the global scope to provide health services for patients, has never faced such a massive amounts of electronic data or experienced such a sharp growth rate of data today. As stated by the Institute for Health Technology Transformation (iHT, U.S. health care data alone reached 150 exabytes ( bytes) in 2011 and would soon reach zettabyte ( bytes) scale and even yottabytes ( bytes) in the future [1]. However, if no appropriate technique is developed to find great potential economic values from big healthcare data, these data might not only become meaningless but also require a large amount of space to store and manage. Over the past two decades, the miraculous evolution of data mining technique has imposed a major impact on the revolution of human's lifestyle by predicting behaviors and future trends on everything, which can convert stored data into meaningful information. These techniques are well suitable for providing decision support in the healthcare setting. To speed up the diagnosis time and improve the diagnosis accuracy, a new system in healthcare industry should be workable to provide a much cheaper and faster way for diagnosis. Clinical decision support system (CDSS), with various data mining techniques being applied to assist physicians in diagnosing patient diseases with similar symptoms, has received a great attention recently [2]– [4]. Naïve Bayesian classifier, one of the popular machine learning tools, has been widely used recently to predict various diseases in CDSS [27]. Despite its simplicity, it is more appropriate for medical diagnosis in healthcare than some sophisticated techniques [6], [7].