Big data analytics in intensive care units: challenges and applicability in an Argentinian hospital

Abstract

In a typical intensive care unit of a healthcare facilities, many sensors are connected to patients to measure high frequency physiological data. Currently, measurements are registered from time to time, possibly every hour. With this data lost, we are losing many opportunities to discover new patterns in vital signs that could lead to earlier detection of pathologies. The early detection of pathologies gives physicians the ability to plan and begin treatments sooner or potentially stop the progression of a condition, possibly reducing mortality and costs. The data generated by medical equipment are a Big Data problem with near real-time restrictions for processing medical algorithms designed to predict pathologies. This type of system is known as realtime big data analytics systems. This paper analyses if proposed system architectures can be applied in the Francisco Lopez Lima Hospital (FLLH), an Argentinian hospital with relatively high financial constraints. Taking into account this limitation, we describe a possible architectural approach for the FLLH, a mix of a local computing system at FLLH and a public cloud computing platform. We believe this work may be useful to promote the research and development of such systems in intensive care units of hospitals with similar characteristics to the FLLH.