Transcript MDPaedigree NND AmsterdamNL 23feb2015 – WP6WP11

M24 meeting MP-Paedigree
Amsterdam, 23 February, 2015
Neurological and Neuromuscular Diseases
WP6 – WP11
MD Paedigree for NND
MODEL DRIVEN, 2 tracks
1. Individualized musculoskeletal models to enhance gait analysis
as an etiological focused diagnostic tool
2. To employ pattern recognition statistical tools, on large
datasets, to enable prediction of treatment outcome
NND M24 meeting / February 2015
Main tasks after 1st year review by EU
1. WP6 : speed up data collection
2. WP 11: clarify role of DXA
3. Suggest adaptation of DOW
4. Update self assessment
After M12 self-evalution:
5. Intensify contact infrastructure group
NND M24 meeting / February 2015
T6.1 – QA on data collection and clinical
protocols
D6.1 : CGA clinical protocols at M18
Status at M18:
Task T6.1 .1 Technical Quality Assurance (TQA) on clinical protocols and on data collection
• TQA and data collection protocols agreed
• TQ assessed accordingly in all 3 centres (processing finalized
by M24 > D6.2)
Task T6.1 .2 Definition of standard clinical protocol for CGA
• Standard and extended protocols agreed
• D6.1 Delivered @ M18
T6.1 – QA on data collection and clinical
protocols
• D6.2 : sharing CGA clinical protocols at M24
• Status at M18:
Ready for sharing at ESMAC 2014 in Rome
• Status at M24 : to be reported today
T6.2 – Gait analysis collection CP
D6.3 : report on collected data at M36
Status at M18:
Task T6.2.1 Retrospective data for CP
OPBG : 200/200 identified 50 ready for upload
Leuven: 100*/400 identified 25* ready for upload
(20 / month to go)
*) mixed pre & post treatment
DOW changes: minor
Status at M24 : to be reported today
T6.2 – Gait analysis collection CP
D6.3 : report on collected data at M36
Status at M18:
Task T6.2.1 Retrospective data for CP
OPBG : 200/200 identified 50 ready for upload
Leuven: 100*/400 identified 25* ready for upload
(20 / month to go)
*) mixed pre & post treatment
DOW changes: minor
Status at M24 : to be reported today
T6.2 – Gait analysis collection - CP
D6.3 : report on collected data at M36
Status at M18:
Task T6.2.2 Prospective data for CP
• Protocols agreed
• Ethical approval reached
• Age range extended
• First patients measured (OPBG)
• KUL and Vumc to start (1 per 1,5 month)
• DOW changes none
• Status at M24 : to be reported today
T6.3 – Gait analysis collection – DMD/CMT
D6.3 : report on collected data at M36
Status at M18:
Task T6.3 Prospective data for DMD & CMT
• Protocols agreed
• Ethical approval pending or reached
• Age range extended, first DMD patients measured
• DOW changes: 10 DMD, 10 CMT, 2 times CGA per centre
(KUL OPBG) 1 time an MRI
• Status at M24 : to be reported today
T6.4 – Image acquisition
D6.4 : report on collected data at M44
Status at M18:
• Protocols agreed
• Ethical approval reached in all centres
• Age range extended, CMT1 replaces SMA
• OPBG : 14 MRI sets healthy children ,
11 patients: 7 DMD, 1 CMT1A, 3 CP
VUmc : 1 healthy adult (test scan)
KULeuven : starting
• DOW adjustments: no DXA scans (abandoned, see WP11)
• Status at M24: to be reported today
NND - WP11
DOW adjustments and next steps
MD-P M24meeting - Amsterdam
Developments in WP11
Analysis after meeting July 2014
• No more DXA scans and analysis
 Too little evidence for added value DXA
 Problems with ethical approval
• Less emphasis on mass distribution model
 Not sensitive to model output
 Mass and inertia tensor still calculated using
subject-specific bone/muscle/fat volumes; but no
bone mass distribution (as could be based on DXA)
• Steps from MRI to model more complex than expected
(see red circles scheme next slide)
Main changes needed in WP11
• More emphasis on sensitive parameters
 Based on sensitivity study
I.e. joint centers and axes; muscle attachments;
muscle/tendon lengths
 more effort is needed to make this happen
USFD can play a role in this instead of DXA
image processing
Details of WP11 changes needed
(see next sheet)
• Wr.t. Joint centers / axes
 Comparison:
 Based on bone meshes and MRI markers (USFD)
 Based on gait data (MM)
 Based on combination of both (USFD)
 Advanced methods explored to create subjectspecific joint models (not necessarily hinge/ball)
using bone surface meshes (USFD)
Details of WP11 changes needed (2)
(see previous sheet)
• W.r.t. muscle attachments
 Not easily identifyable from MRI
 New methods explored to follow muscle belly
midline to bone mesh (USFD)
 + Can possibly be morphed using atlas data?
(Siemens)
Details of WP11 changes needed (3)
• W.r.t. muscle fiber / tendon slack lengths
 Idea to use ultrasound measurements for this,
but seems not a (realistic) option within this
project anymore
Next steps
• Siemens
 Try out Leuven annotated MRI images as atlas
 Use this to create annotated MRIs of our own data
(incl bony landmarks and attachment points)
• Sheffield
 Create accurate joint models based on meshes
 Create methods for muscle attachment site (using
mesh morphing or extrapolation of muscles)
 Conversion from annotated-MRI to model
parameters
Next steps
• TU Delft
 Create patient-specific OpenSim models with USFD
input
 Sensitivity / effect studies of model changes
 Validation against measured data
 Explore predictive models
• Motek
 Create workflow for personalization without
imaging
Using functional joint centers; accurate length scaling;
bone deformities; …
 Apply TUD image-based models in HBM
SYSTEM INTEGRATION OVERVIEW
MD-PAEDIGREE WP11
MRI data OPBG
(30)
MRI data KUL (30)
MRI data VUMC
(10)
1A
1B
SIEMENS
2
USFD
3
TUD
1C
4A
4B
4C
GAIT+PE data
OPBG
GAIT+PE data KUL
GAIT+PE data
VUMC
6
5A
MOTEK
5B
5C
Brief description of output – input data
1. MRI images (Dicom format; stacks combined to one image)
2. Segmented MRI images (… format; bones and muscles segmented; anatomical landmarks & attachment sites indicated)
3. Model parameters from MRI (… format; muscle volumes; segment mass & inertia; femur and tibia torsion; joint centers/models; attachment
sitesl; in segment reference frames)
4. Gait and calibration trials (C3D/MOX format – same as 5) + (instrumented) physical exam; dynamometry; O2; anamnesis (txt/pdf format)
5. Gait and calibration trials (C3D/MOX format – same as 4)
6. Personalized OpenSim models (.osim format)
7. HBM output to repository
7