MICCAI grid tutorial
• Overview
– IXI project summary – Derek Hill – Wrapping image analysis algorithms as grid services - Daniel Rueckert – break – Example medical imaging applications using wrapped services – Derek Hill & Daniel Rueckert
KCL Imperial Oxford http://www.ixi.org.uk
Information eXtraction from Images (IXI)
• 3 year e-science project funded by core programme
– Additional support from GSK, Philips Medical Systems, medical charities
• Uses grid-enabled image registration and segmentation for drug discovery, medical research, and decision support in healthcare.
The IXI consortium
• King’s College London (Derek Hill, Dave Hawkes, Steve Williams, Gareth Barker) • Imperial College (Jo Hajnal, Daniel Rueckert, David Edwards, LESC) • University of Oxford (Steve Smith) • Jointly coordinated by Derek Hill & Jo Hajnal • 9 full time researchers
Why IXI?
• We call this project Information eXtraction from Images to emphasize the key concept which is using image analysis to generate image metadata – information about the images – and the generic applicability of this technology.
IXI components
• Grid toolkit (algorithms wrapped as grid services) for medical image analysis • Workflow language and engine for execution • Distributed data and researchers
Compute-intensive application scenarios
• Rheumatoid Arthritis with GSK quantification of change • Facilitating international collaboration on algorithm validation. • Dynamic brain atlas for decision support • Brain development of neonates All involve:
– large scale analysis, – aggregation of data, – simple end-user interface to the grid infrastructure.
Automatic registration technology wrapped as grid services
Rigid registration of MR and CT images of the head
Inter-subject image warping
Identifying differences between term-born and premature babies
• High dimensional non-rigid registration • Combined PCA and LDA • Pilot study
– Premature neonates scanned at term equivalent age – Both signal changes and morphometry changes
• 97% classification efficiency + visualise differences
Structural difference
Signal difference
Drug discovery: automatic quantification of change in joints
• Workflow:
• • Rigid registration to find corresonding bones (top movie) Nonrigid registration: warp one image into another one to quantify subtle change (bottom movie) Grid infrastructure provide distributed computing and provenance for high throughput analysis
•
Collaborations
• • • • Neurogrid (MRC funded) Steve Jarvis (Warwick) IBIM (EPSRC LSI) INRIA and CREATIS in France
– Healthgrid demo
• Chicago/Argonne NL in USA
Next presentation
• The grid for collaborations in image analysis research • Wrapping image analysis algorithms as grid services (GT3) and incorporating into workflows (Daniel Rueckert) • Incorporating image analysis algorithms into the Virtual Data System / GT2
– MRI brain atlas generation – Delineation of joints from dynamic MRI