Biomedical Engineering Reference
In-Depth Information
•
Export of CellML files to various programming (e.g. C/C++, F77, Java MAT-
LAB, Python) and word processing languages (MS Word 2007/2010 and T
E
X).
•
Import/export from/to SBML.
Simulating files
•
Editing of simulation (e.g. start/end points) and numerical solver parameters (de-
pendent on the chosen solver);
•
editing of model parameters (using a tree-like view);
•
support for models consisting of DAEs (using CVODE and IDA as default solvers);
•
run/pause/stop a simulation;
•
plotting of simulation results (against any model parameter);
•
export of simulation results to a comma-separated value format;
•
create a new or update an existing CellML file based on the results of a simulation.
Analysis features will mainly be provided by the community through the use of plug-
ins (e.g. a plugin to analyse cardiac action potentials and extract some key parameters
from them). The anticipated public release date for OpenCOR, at which point it will
replace the existing OpenCell software, is December 2011.
8.6 FieldML
To cater for models that do include spatial information, another markup language
called FieldML (fieldml.org) is being developed. FieldML files contain all the pa-
rameters and expressions needed to mathematically define fields over multidimen-
sional manifolds representing space, time and other domains. The most common
form of parameterization is a finite element mesh, where the representation is built
out of nodal parameters, mesh topology and the element basis functions that inter-
polate the nodal parameters over the elements. However, FieldML is intended to
handle more general parameterizations than just finite element fields and also in-
cludes dense data formats (e.g. for images embedded inside models) and arbitrary
functions of existing fields.
FieldML is being developed as a standard for communicating computational field
descriptions and data (Christie et al., 2009). The fundamental idea behind its design
is to define the spaces that make up a body of interest (the domain - a manifold)
from basic primitives, and to map the positions on those spaces to values (the fields)
using explicit mathematical expressions.
Important principles in its design include:
•
use of a minimal set of interoperable concepts;
•
placing no limits on the generality of the domain (discrete, continuous or any
combination; representing space, time, parameter-space, population, etc. or any
combination) so that all quantities are 'field-like';
•
having no 'low-level objects', only domains/sets with attributes mapped to them;
•
extensibility to support arbitrarily complex field functions;
•
supporting reuse of domain descriptions and functional expressions;
