BiosignalML

Introduction

The Physiome Project has established two XML standards for encoding models of physiological processes over the past 10 years. CellML (www.cellml.org) is a standard for encoding spatially lumped models, such as the systems of ODEs and nonlinear algebraic equations that characterize cellular processes. FieldML (www.fieldml.org) is an XML standard encoding parameterized spatial fields, such as finite element models of mammalian anatomy and the spatially varying fields that result from the solution of systems of partial differential equations used to characterize organ level function. There is now a need to establish data standards that can be used for encoding the input data for models based on CellML and FieldML, or for the output of these models. Note that there is currently a standard for image data (DICOM) but not for time-varying biomedical signals.

We therefore propose the development of a new standard for the encoding and storing of biomedical time-varying signals such as the electrocardiogram (ECG), electromyogram (EMG), electroencephalogram (EEG), electrocorticogram (ECoG), electrooculogram (EOG), blood pressure and signals associated with respiration, speech, etc. The BiosignalML standard will need to include metadata annotation that provides information on the context for the recorded signal, using community accepted bio-ontologies where appropriate. The standard will also need to include dense data formats for efficient storage of large data sets.

Along with the development of this BiosignalML standard, we propose to establish a web-accessible repository of signals encoded in the new standard and to develop an open source software toolkit for displaying, processing, annotating and analyzing these signals. The signal repository would be used by CellML and FieldML model simulation software via a BiosignalML API. The BiosignalML toolbox will need to include a broad spectrum of techniques in common use, such as FFT, wavelet, spectral, BSS, entropy methods, statistical methods, etc. Tutorials will need to be developed to illustrate use of the BiosignalML standard and associated software.

The Physiome models would be used to link signal processing based on BiosignalML from multiple spatial and temporal scales. Standards would also be established for lead placement in relation to the Physiome anatomical models and the ability to display time dependent spatial maps from multi-electrode recordings on the various anatomical models. Note that the Physiome Project will be establishing a database of parameterised body shapes with tools to customise whole body models (or organs) for individuals based on limited body shape measurements.

Current related efforts

Physionet (www.physionet.org) is a fairly comprehensive resource for physiological signals (especially neural). It includes a repository of signals and an open source (GPL) toolkit.

The BioSig project (biosig.sourceforge.net) is an open source software library for biomedical signal processing with a GUI built on Qt and close links to Matlab and Octave.

Note that the goals of the Physiome BioSignalML project are much broader than these two projects but it will clearly be important to develop links with these groups.

There are already some projects underway to establish XML standards for time series, including one by the FDA, and another called SleepXML (www.sleepxml.net).

Participants

The following people have expressed an interest in being involved in this project:

• Prof Sergio Cerutti 
  Department of Bioengineering, Polytechnic University of Milan, Italy
  This e-mail address is being protected from spambots. You need JavaScript enabled to view it.
• Prof Forbes Dewey
  Dept of Mechanical Engineering, MIT, US
  This e-mail address is being protected from spambots. You need JavaScript enabled to view it.  
• Prof Rod Hose
  Dept of Biomedical Engineering, University of Sheffield, UK
  This e-mail address is being protected from spambots. You need JavaScript enabled to view it.  
• Prof Nigel Lovell
  Graduate School of Biomedical Engineering, Univ of New South Wales, Sydney, Australia
  This e-mail address is being protected from spambots. You need JavaScript enabled to view it.    
• Dr Grace Peng
  National Institute of Biomedical Imaging & Bioengineering
  Bethesda, MD, USA
  This e-mail address is being protected from spambots. You need JavaScript enabled to view it.  
• Dr Mark Titchener, Mr David Brooks, A/Prof Bruce Smaill, Prof Peter Hunter
  Bioengineering Institute, University of Auckland, New Zealand
  This e-mail address is being protected from spambots. You need JavaScript enabled to view it. ,  This e-mail address is being protected from spambots. You need JavaScript enabled to view it. ,  This e-mail address is being protected from spambots. You need JavaScript enabled to view it. , This e-mail address is being protected from spambots. You need JavaScript enabled to view it.