Biomedical Engineering Reference
In-Depth Information
Fig. 8.8.
Integration of the Taverna workflow manager within GIMIAS allowing visual design of
complex workflows and the option of local or remote execution of each module. On the left the
definition of a single segmentation workflow from cerebrovascular 3dRA images, on the right the
resulting surface mesh visualized in GIMIAS
The user can drag and drop the available services to the workflow diagram area
and connect the inputs and outputs of the used services. Parameter sweep and basic
conditional flow control parameters can be used, specifying, for example, ranges for
input parameters or the execution of a certain pipeline on a specified type of result.
When the workflow is executed, the input data for the workflow is automatically
passed from GIMIAS to Taverna. When the service being executed is located re-
motely, the data is transferred using SSH protocol between local and remote com-
puters. Once the workflow is finished, the output data will be automatically loaded
in GIMIAS.
For example to create the simple segmentation workflow described in Fig. 8.8,
the following processing services have been used: ITK Binary threshold, ITK Gaus-
sian blur and VTK marching cubes. During execution, the input image specified in
GIMIAS is processed and the resulting surface mesh is automatically rendered in
the GIMIAS visualisation environment.
8.9 SED-ML
15
The Simulation Experiment Description Markup Language (SED-ML) is an XML-
based format for encoding simulation experiments, following the requirements de-
fined in the MIASE guidelines (biomodels.net/miase). SED-ML specifies which
model to use, the experimental task to run and which result to produce.
SED-ML is built of five main classes: the Model Class, the Simulation Class, the
Task Class, the DataGenerator Class, and the Output Class.
The
Model class
is used to reference the models used in the simulation experi-
ment. SED-ML itself is independent of the model encoding underlying the models.
15
Comments here are modified versions of comments on the SED-ML website sed-ml.org
