International Journal of Computer and Information Technology (ISSN: 2279 – 0764) Volume 03 – Issue 02, March 2014 www.ijcit.com 299 Health Care Infrastructure Security using Bimodal Biometrics System Omobayo A Esan, Selema Ngwira Computer System Engineering Tshwane University of Technology, TUT Pretoria, South Africa {esanoa, ngwirasm}@tut.ac.za Tranos Zuva, Lerato Masupha Computer System Engineering Tshwane University of Technology, TUT Pretoria, South Africa {zuvaT, masuphaLE}@tut.ac.za Abstract— Ensuring the security of patient medical record has become an important issue with this current l electronic medical record, this is due to the fact that patient confidential health information can be access, view and edit on Personal Computer (PC) at any time. In order to protect the patient health confidential information, this research use bimodal biometric (face and fingerprint) authentication to secure patient Electronic Medical Record (EMR). The bimodal biometric (fingerprint and face) are often affected by distortions which are caused by environmental noise such as oil, wrinkles, dry skin and dirt. These often affect the biometric system accuracy during authentication stage. In order to protect and improve the accuracy of patient Electronic medical Record (EMR) in health care infrastructure, this study introduced Modified Gabor Filter (MGF), a fast Principal Component Analysis (PCA) algorithm with Support Vector Machine (SVM) to address the issues of fingerprint and face image distortion respectively. From the experiment conducted from 1000 patients with 20 fingerprint and face image from each to give 50 test cases. The result shows that the proposed bimodal biometrics approach gives a lower False Rejection Rate (FRR) and False Acceptance Rate (FAR) and this shows better constructive accuracy of our system on patient EMR in health care infrastructure. I. INTRODUCTION An automated technique of recognizing a person based on physiological and behavioural traits is known as a biometrics system. The physiological traits include the face, fingerprint, palm print and iris, which remain permanent throughout an individual’s lifetime. The behavioural traits are signature, gait, speech and keystroke, etc., which change over time [1]. The advantages of a fingerprint authentication system make the system the most widely used biometric system for various applications for security and access control in airports, at borders, immigration offices, houses, offices, banks and other places where security needs to be enhanced [1],[2]. However, face identification is also one of the acceptable biometric systems widely used in public security systems, attendance systems etc. because of its convenience and high efficiency [1],[2]. In this regards, the problem of securing information emerged, since information needs to be managed and secured data [1]. However, in this present technological age securing of Electronic Medical Records is becoming an increasingly important problem due to the fact that most health professionals can edit and view a patient’s medical record using a tablet PC at any point of time across the network [3]. Thus, this has subjected patient’s medical information to lack of privacy and confidentiality. It is therefore of high importance to address this issue of unsecured patient medical record released to health practitioners without the permission of authorized patient [4]. According to the study conducted by High bit Security in 2011on patient privacy and data breach security in United State of America, the investigations shows that 94% of hospital survey suffers data breaches and this cost US about $7 billion annually. Fig. 1 shows the record of security breaches at health care institutions from 2008 to 2010 [5]. Fig. 1. Visualization of record of security breaches in High bit Security