Session S43.3

An XML Format for Storing Body Surface Potential Map Recordings

R Bond*, D Finlay, C Nugent, G Moore

University of Ulster
Belfast, UK

A format for storing electrode positions and Body Surface Potential Map (BSPM) ECG data have been developed within this study using the eXtensible Markup Language (XML) Formats such as DICOM aim to support all diagnostic mediums ranging from radiographic images to the ECG. Nevertheless, there are a range of ECG specific formats such as SCP-ECG, aECG, ecgML and XML-ECG. These recognized formats mainly support the 12-lead ECG and lack support for less prominent ECG acquisition methods. This has lead to the development of new ECG storage formats that support a particular ECG acquisition method, e.g. Holter data has been stored using XML. Recognizing this trend towards ECG acquisition specific solutions based on XML, a similar solution for storing a BSPM has been developed here. The 12-lead ECG utilizes 10 electrodes placed at well defined anatomical landmarks. BSPMs on the other hand can utilize over 200 electrodes placed at custom anatomical landmarks and there is no standard electrode configuration. The format developed in this study not only facilitates the storage of the BSPM data, but also stores the electrode layout. This is useful for both clinical reference and subsequent representation of the data e.g. contour plotting. In this study Scalable Vector Graphics (SVG) were used to allow the description of electrode locations. SVG is an XML language for describing images allowing them to be scaled better and stored more efficiently in comparison to rasterized equivalents. The format stores each BSPM recording as one XML file. An SVG torso schematic is stored within the header of the file and the ECG lead data is stored in the body. The data from each lead is labeled with a set of X, Y coordinates that relate to the SVG torso diagram and hence defining the position of that lead. To illustrate this format, a Lux-192 BSPM recording was stored. The lead layout and an averaged beat from each of the 192 leads, sampled at 1000Hz, resulted in a 291 kilobyte XML file. This file was subsequently reduced in size to 76 kilobytes using a standard compression algorithm. In this study we have demonstrated how XML can be used to allow robust and compact storage of BSPM data.

(Abstract Control Number: 228)