Biomedical Engineering Reference
In-Depth Information
•
Gene Regulation
•
Myofilament Mechanics
•
Hepatology
•
Neurobiology
•
Immunology
•
pH Regulation
•
Ion Transport
•
PKPD
•
Mechanical Constitutive Laws
•
Signal Transduction
•
Metabolism
•
Synthetic Biology
These categories change as new classes of models are developed. The 'PKPD'
(PharmacoKinetic-PharmacoDynamic) classification, for example, has recently been
added to accommodate the models being developed by this community and encoded
in CellML.
8.12 OpenCMISS
16
We discuss one of several software packages being developed for multiscale physi-
ological simulation of partial differential equations (PDEs). Others are Chaste (com-
lab.ox.ac.uk/chaste) and Continuity (cmrg.ucsd.edu/Continuity).
Background to CMISS
CMISS is an acronym for 'Continuum Mechanics, Image analysis, System identifi-
cation and Signal processing'. Code development for CMISS began in Auckland in
1980 with the goal of creating a bioengineering finite element code for solving mul-
tiple coupled partial differential equations. The initial applications were for the heart
physiome project, primarily for electro-mechanics (Hunter & Smaill, 1989; Hunter,
Pullan and Smaill, 2003), but it rapidly expanded to support applications in other or-
gans such as the lungs, with coupling between soft tissue mechanics, heat and mois-
ture transport and airway fluid mechanics (Tawhai et al., 2000, 2004). While finite
element methods provided the primary numerical technique, the code was further
developed in the 1980s to include boundary element methods, for electrical current
flow in the thorax (Bradley et al., 1997), and finite difference equations based on
curvilinear grids.
Some of the unique features of CMISS (in comparison to commercial finite ele-
ment codes) that have been carried over into OpenCMISS are: (i) the use of cubic
Hermite basis functions to preserve
C
1
or
G
1
continuity across element boundaries
where this is important for efficient field representation (Bradley et al., 1997); (ii)
the use of variable order basis functions (e.g. bicubic Hermite by linear Lagrange);
and (iii) the close relationship that is preserved between the representation of tissue
structure and the organ anatomy, through the use of material structure fields defined
with respect to embedded material coordinates.
The CMISS code (including cmgui) has been at the heart of many multi-scale
physiome projects over the past 20 years, but as the VPH/Physiome modelling stan-
dards evolved it became clear that a redevelopment of both programs was needed to
16
A more extensive overview of OpenCMISS, on which this section is based, is given in Bradley
et al., 2011.
