Real Time Data Analytics Architecture for ECG

The concept of big data emerging from the expansion of data volumes on the Internet has begun to talk about its name in medicine as well as in many fields of life. Big data analytics, which also require the use of machine learning methods, enable the use of decision-making processes by extracting useful information from large and complex data sets. Implementing machine learning strategies on data sets within big data is an expensive process because it requires extensive use of resources such as CPU and memory. For this reason, platforms specially developed for big data analytics are designed. One of these systems, Apache Spark, has built-in machine learning algorithms ranging from regression to classification and clustering, and is a very powerful engine for real time stream processing. In this study, the first results of a system that provides real-time disease diagnosis from ECG data using Logistic Regression are presented. The first findings obtained show that this architecture, built with Apache Kafka and Apache Spark, can be an important design option in real time processing of ECG data.