摘要：In a typical intensive care unit of a healthcare facilities,many sensors are connected to patients to measure highfrequency physiological data. Currently, measurements areregistered from time to time, possibly every hour. With thisdata lost, we are losing many opportunities to discover newpatterns in vital signs that could lead to earlier detectionof pathologies. The early detection of pathologies givesphysicians the ability to plan and begin treatments sooneror potentially stop the progression of a condition, possiblyreducing mortality and costs. The data generated by medicalequipment are a Big Data problem with near real-timerestrictions for processing medical algorithms designed topredict pathologies. This type of system is known as realtimebig data analytics systems. This paper analyses ifproposed system architectures can be applied in the FranciscoLopez Lima Hospital (FLLH), an Argentinian hospitalwith relatively high financial constraints. Taking intoaccount this limitation, we describe a possible architecturalapproach for the FLLH, a mix of a local computing systemat FLLH and a public cloud computing platform. We believethis work may be useful to promote the research anddevelopment of such systems in intensive care units of hospitalswith similar characteristics to the FLLH.
关键词：Clinical Decision Support System; Real Time;Big Data; Cloud Computing; Intensive Care Units
