Real-time custom processing and delivery of large ECG data sets

A tool for rapid custom-processing large data sets of ECC signals has always been in demand in research communities. To facilitate this requirement, a link must be established between the numerical processing engine and the data storage system. In our research, we have established an extensive collection of ECG signals, known as the Glasgow Data, which is accessible to subscribers over the Internet (see related Computers in Cardiology paper ´Implementation of Web Database for ECG, M.A. Paracha, S.N. Mohammad, P.W. Macfarlane, J.M. Jenkins´). Meaningful analysis, such as specific plots and images of patient ECG signals, are stored as images in this large database. Initially, these images are generated in a numerical processing and scientific visualization tool such as Matlab. The key to rapid deployment of results from these processing tools is the facility directly, in real-time, to provide results to online users of the database. This requires connectivity between the processing engine and the database back-end. In this design we use open-source modules to establish this link, and demonstrate usability in the field of electrocardiology. With over 3300 individual patients registered in the Macfarlane Glasgow database, we are able to run researcher-specific algorithms over the entire domain of signal data and provide results instantly over the web-based system. This allows researchers to mold and study the data according to their own requirements and thus transform the data into useful knowledge, which should play a vital role in their respective research.