ECGlab: User friendly ECG/VCG analysis tool for research environments

We present ECGlab, a cross-platform, user friendly, graphical user interface for assessing results from automated analysis of ECGs in research environments. ECGlab allows visual inspection and adjudication of ECGs. It is part of our recently developed framework to automatically analyze ECGs from clinical studies, including those in the US Food and Drug Administration (FDA) ECG Warehouse. ECGlab is written in C++ using open-source libraries. Supported ECG formats include Physionet, ISHNE and FDA XML HL7. ECG processing and automated analysis is done with ECGlib (ECG analysis library). ECGs can be loaded individually or grouped using ECGlib index format and information such as demographics or signal quality metrics can be loaded from metafiles to navigate through the ECGs and guide their review. The user can graphically adjudicate the ECGs in a semi-automatic or manual fashion. Vectorcardiograms can be assessed as well. A prototype for automatic extraction, based on heart rate stability and signal quality, of 10 seconds ECGs from continuous Holter recordings is also available. ECGlab, which has been successfully tested in Linux and Microsoft Windows, is currently being used to assess ECGs from clinical studies. We are working on making ECGlab open-source in order to facilitate ECG research.