Ulster University Logo

Ulster Institutional Repository

A compressed xml format and a rich internet application for procesing and visualizign body surface potential maps

Biomedical Sciences Research Institute Computer Science Research Institute Environmental Sciences Research Institute Nanotechnology & Advanced Materials Research Institute

Bond, R, Finlay, D, Nugent, CD and Moore, G (2009) A compressed xml format and a rich internet application for procesing and visualizign body surface potential maps. In: The Northern Ireland Biomedical Engineering Society, Belfast. UNSPECIFIED. 1 pp. [Conference contribution]

Full text not available from this repository.

Abstract

The 12-lead ECG remains the gold standard for detecting cardiac pathologies such as Myocardial Infarction (MI). Nevertheless research suggests that analyzing the Body Surface Potential Map (BSPM) improves the accuracy in diagnosing not only MI but Left Ventricular Hypertrophy, Cardiomyopathhy and Wolf-Parkinson-White syndrome [1]. The 12-lead ECG incorporates 10 electrodes located at standard positions on the body whereas BSPMs have no standard number of electrodes. In fact BSPMs have utilized between 32 and 213 electrodes located at various anatomical landmarks on the torso [2].Recently, numerous formats for storing ECG data have been established such as SCP-ECG, DICOM, ECGml, and HL7 aECG [3-5]. These formats concentrate on storing 12-lead ECG data and therefore have no requirement to store anatomical electrode positions in the data file since 12-lead electrode positions are standardized. These motives lead to the development of an XML format within the current work for storing BSPM data since BSPMs unlike the 12-lead have no standard anatomical electrode arrangements [2]. Therefore this XML BSPM format can store the anatomical landmark of each electrode. This concept has been achieved by storing an unrolled torso schematic in the header of the ECG data file using Scalable Vector Graphics (SVG) [6]. The SVG markup is used as a reference diagram and each lead locates its anatomical position using an X and Y 2D coordinate system. As a result this format supports all BSPM data and electrode layouts. The reason for choosing XML and not a binary format is that XML has a family of related technologies such as SVG, XSLT and Schema to name but a few. Another reason is the practicality of influencing predominate standards such as DICOM, HL7 and SCP-ECG. Therefore XML allows such standards to easily integrate conceptual techniques from this BSPM format. On the other hand, XML verbosity can be an issue. To address this we compressed the format using the DEFLATE algorithm hence reducing the file size by 74%. In order to evaluate this format it was essential to develop ECG viewing software [9]. For development we choose Adobe Flash due to its lossless vector quality, since ECG viewers historically visualized waveforms by coloring pixels [7] [8]. The user uploads a BSPM file and the viewer using ActionScript 3.0 decompresses the ZIP file, parses the XML, renders the SVG diagram, and draws the waveforms using vector graphics and all this happens within the user’s browser. From here users can investigate each waveform in a comprehensive user interface using a wide range of tools, including callipers. The XML BSPM format has an optional ‘<transformations>’ element where equations are stored to transform the BSPM into for example the 12-lead ECG. If this element exists the application displays the 12-lead transformation button which when clicked executes the equations and displays the derived 12-lead ECG either as waveforms or as an interactive 3D torso. Using this new format and the Rich Internet Application, experts can process, visualize and scrutinize their custom lead sets, which in return can assist in BSPM research.

Item Type:Conference contribution (Paper)
Faculties and Schools:Faculty of Computing & Engineering
Faculty of Computing & Engineering > School of Computing and Mathematics
Research Institutes and Groups:Computer Science Research Institute
Computer Science Research Institute > Smart Environments
ID Code:7070
Deposited By:Professor Christopher Nugent
Deposited On:29 Jan 2010 14:18
Last Modified:25 Jan 2013 14:36

Repository Staff Only: item control page