Description
This section includes some medical imaging community contributions to the exploitation of the EGEE grid infrastructure. Applications to cardiac, brain and lung image analysis are presented. Emphasis is put on the domain services deployed on top of the core middleware to implement the applications and conduct experiments. In particular, workflow tools are extensively used to describe applications logic and provide a high level interface to the grid. Application-level reliability improvement mechanisms are also being described
-
Mr Johan Montagnat (University of Nice Sophia-Antipolis - I3S)13/04/2010, 16:00Scientific results obtained using distributed computing technologiesOralWe will present the description of a cardiac image analysis workflow with the new Gwendia language and its enactment on two different grid platforms, namely EGEE (using the MOTEUR workflow engine) and Grid'5000 (using the DIET workflow engine). Technical achievements enabling the execution of the very same workflow on those two grids will be presented. Finally, early application results (both...
-
Mr Tristan Glatard (CNRS - CREATIS)13/04/2010, 16:20Experiences from application porting and deploymentOralThis presentation summarizes experience acquired at Creatis with some Scientific Gateway components to deploy medical imaging applications on the EGEE grid during the last years. We will detail feedback about various tools that we tested, including pilot-job systems, workflow managers, user front-ends, monitoring components and data management systems. Based on application use-cases ported at...
-
Dr Silvia Olabarriaga (Academic Medical Center of the University of Amsterdam)13/04/2010, 16:40Experiences from application porting and deploymentOralMedical image analysis requires large computational effort. The construction of “atlases” specific for each study of a specific brain disease is a good illustration: typically all images need to be aligned (registered) to each other. In this work we explore the Dutch Grid (part of EGEE) to implement a service that automates the construction of such atlases for Diffusion Tensor Imaging (DTI),...
-
Mr Ihar Safonau (United Institute of Informatics Problems of the National Academy of Sciences of Belarus, Minsk, Belarus)13/04/2010, 17:00Scientific results obtained using distributed computing technologiesOralThe problem of indexing and content-based retrieval of medical images in a distributed computing infrastructure is discussed in the paper. It is considered in the context of a national-wide lung screening and diagnosis programme which is currently under development. High-resolution images are drawn from a test database containing results of X-ray chest examination of about 250000 people. The...