Rao et al Journal of Drug Delivery & Therapeutics. 2019; 9(1-s):358-360 ISSN: 2250-1177 [358] CODEN (USA): JDDTAO Available online on 21.02.2019 at http://jddtonline.info Journal of Drug Delivery and Therapeutics Open Access to Pharmaceutical and Medical Research © 2011-18, publisher and licensee JDDT, This is an Open Access article which permits unrestricted non-commercial use, provided the original work is properly cited Open Access Research Article Fast Pattern Discovery in Healthcare Data Using Graphics Processors Naseem Rao, Safdar Tanweer* Assistant Professors, CSE Department, Hamdard University, Delhi, India ABSTRACT The mobile medical diagnosis and health monitoring system helps in managing the various chronic diseases like asthma, blood pressure and heart diseases etc. in consultation with the remotely available physicians by initiating the emergency call automatically on the physician’s mobile phone and providing the on-line vital medical parameters captured by the body area sensor network of the patient. We observed that a GPU based solution can outperform a CPU based solution by more than 30% in terms of speed up, while giving same accuracy of results, divided among healthy, normal and unhealthy patients. Finally, key parameter to model our health care data likestandard deviations of {1, 0.5, 0.5}, means of {(1, 1), (0, 0), (-1,-1)} are used to study healthy persons and unhealthy patients. Keywords: Healthcare ; GPU; EEG; PCG; datastructure Article Info: Received 11 Jan 2019; Review Completed 19 Feb 2019; Accepted 19 Feb 2019; Available online 21 Feb 2019 Cite this article as: Rao N, Tanweer S, Fast Pattern Discovery in Healthcare Data Using Graphics Processors , Journal of Drug Delivery and Therapeutics. 2019; 9(1-s):358-360 DOI: http://dx.doi.org/10.22270/jddt.v9i1-s.2446 *Address for Correspondence: Safdar Tanweer, Assistant Professors, CSE Deptt., Hamdard University, Delhi, India INTRODUCTION Today, we are witnessing a rapid surge in amount of data growth. Data acquisition methods are also improving day-by- day. The embedded patient monitoring systems also generate huge amount of data. Therefore, one of the goals of this research is to understand how this health care related data of patients can be used to come up with better solutions. Several aspects of the developed system related to design, usability, reliability, and accuracy etc. will also be studied in actual field conditions [1-5]. SIMULATION DETAILS & RESULTS Our algorithm runs on a CPU-GPU hybrid system. The parallel algorithm was implemented on NVIDIA GPU using CUDA-C. Not all part can be ported to GPU as discussed above. But the portion of the algorithm which can be ported is a highly compute intensive – calculation of distance. In Figure 1 we show the health care data set consisting of 16K points based on the ground truth. Figure 1: Healthcare data based on ground truth