Information Technology Reference
In-Depth Information
The gUSE framework interfaces directly with the gLite middleware services.
More speci
cally, the DCI Bridge sends grid jobs to the WMS. However, sending a
large number of grid jobs to the WMS is not ef
cient, especially if their run-time is
short. Therefore, many user communities who submit large number of grid jobs
utilize a pilot job framework to optimize the resource utilization, which is a useful
feature to add to the DCI Bridge in the future.
At the time of the development of the AMC-NSG, the data transport between
different DCIs was not supported in a programmatic way. Therefore, we developed
the DT component, which is tailored to the AMC-NSG requirement to transport
data between XNAT (HTTP) and grid data resources (LFC). In the meantime, the
Data Avenue (see Chap. 5 ) has been developed, which will be considered in the
future for the AMC-NSG.
10.4 Conclusions
The AMC-NSG has been available on-line since November 2013 via http://neuro.
ebioscience.amc.nl . The links to its source code, developer, and user documenta-
tions are also available on the SCI-BUS portlet repository [AMC].
Currently, the following medical image processing applications are available to
process Diffusion Tensor Imaging (DTI) and structural Magnetic Resonance
Imaging (MRI) data: (a) Freesurfer: implements segmentation of structural MRI
data with the Freesurfer toolbox (Fischl 2004); (b) DTI-preprocessing: performs
format conversion and quality control of DTI data [DTI]; and (c) BEDPOSTX:
performs local modelling of diffusion parameters with FSL BEDPOSTX [FMRIB].
An overview of SGs for computational neuroscience research is available in
Shahand (2014).
The AMC-NSG has been available to the AMC BIC community since
November 2013. It provides simple web-based user interfaces to reduce the gap
between users, data resources, and DCIs. It also streamlines the data management
by enabling the users to browse the data and metadata that are stored on data
resources, and guiding them to choose the correct data processing application based
on the metadata. Additionally it streamlines the operations of the gateway by
allowing the system administrators to inspect and troubleshoot the problems
transparently, while only exposing high-level error messages to the end-users when
necessary.
At the time of writing, the early
five adopters have successfully processed 700
scans using the AMC-NSG, consuming around 900 days of computing in a
seamless manner. The gateway has also been used successfully in teaching, which
is described in detail in Chap. 16 .
Search WWH ::




Custom Search