MRI analysis for hippocampus segmentation on a distributed infrastructure

Medical image computing raises new challenges due to the scale and the complexity of the required analyses. Medical image databases are currently available to supply clinical diagnosis. For instance, it is possible to provide diagnostic information based on an imaging biomarker comparing a single ca...

Full description

Saved in:
Bibliographic Details
Published in2016 IEEE International Symposium on Medical Measurements and Applications (MeMeA) pp. 1 - 6
Main Authors Tangaro, S., Amoroso, N., Antonacci, M., Boccardi, M., Bocchetta, M., Chincarini, A., Diacono, D., Donvito, G., Errico, R., Frisoni, G. B., Maggipinto, T., Monaco, A., Sensi, F., Tateo, A., Bellotti, R.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.05.2016
Subjects
Biomedical imaging
Distribuited analysis
Hippocampus Segmentation
Loni Pipeline
Magnetic resonance imaging
MRI analysis
Neuroimaging
Pipelines
Protocols
Web services
Workflows
Online AccessGet full text

Cover

More Information
Summary:Medical image computing raises new challenges due to the scale and the complexity of the required analyses. Medical image databases are currently available to supply clinical diagnosis. For instance, it is possible to provide diagnostic information based on an imaging biomarker comparing a single case to the reference group (controls or patients with disease). At the same time many sophisticated and computationally intensive algorithms have been implemented to extract useful information from medical images. Many applications would take great advantage by using scientific workflow technology due to its design, rapid implementation and reuse. However this technology requires a distributed computing infrastructure (such as Grid or Cloud) to be executed efficiently. One of the most used workflow manager for medical image processing is the LONI pipeline (LP), a graphical workbench developed by the Laboratory of Neuro Imaging (http://pipeline.loni.usc.edu). In this article we present a general approach to submit and monitor workflows on distributed infrastructures using LONI Pipeline, including European Grid Infrastructure (EGI) and Torque-based batch farm. In this paper we implemented a complete segmentation pipeline in brain magnetic resonance imaging (MRI). It requires time-consuming and data-intensive processing and for which reducing the computing time is crucial to meet clinical practice constraints. The developed approach is based on web services and can be used for any medical imaging application.
DOI:10.1109/MeMeA.2016.7533716