Österreichischen Gesellschaft für Neurorehabilitation , Graz, 7th November 2014
Structural and functional neuroimaging in neurorehabilitation NICK WARD, UCL INSTITUTE OF NEUROLOGY, QUEEN SQUARE
[email protected] Slides at www.ucl.ac.uk/ion/departments/sobell/Research/NWard or look on
@WardLab
@WardLab
Imaging based neurorehabilitation
Overview 1. Brain imaging in neurorehabilitation – what for? 2. Predicting outcomes and treatment effects after stroke? 3. Biomarkers of plasticity?
Imaging based neurorehabilitation - I Which bit of neurorehabilitation? Rehabilitation is a process of active change by which a person who has become disabled acquires the knowledge and skills needed for optimum physical, psychological and social function
Treatments aimed at reducing impairments
(Task-specific) training cortical stimulation
other
drugs
Imaging based neurorehabilitation - I What type of imaging?
T1 structural
DTI
fMRI
M/EEG
Imaging based neurorehabilitation - I Brain imaging in neurorehabilitation – what for?
1. Predicting long term outcome
e.g. EXPLICIT, PREP, PLORAS
2. Predicting response to treatment
behavioural training plasticity enhancement
Imaging based neurorehabilitation - II Predicting long term outcome after stroke
Imaging based neurorehabilitation - II Predicting long term outcome after stroke - anatomy stroke damage
damaged pathways
cortical reorganisation?
Imaging based neurorehabilitation - II Predicting long term outcome – structural imaging Track from fMRI-defined hand areas in 4 different cortical motor areas
Shultz et al, Stroke 2012
Corrrelation with poststroke hand grip strength
Imaging based neurorehabilitation - II Predicting long term outcome – structural imaging
Imaging based neurorehabilitation - II Predicting long term outcome 1. SAFE = Shoulder Abduction + Finger Extension (MRC scale) 72 h after stroke (range 0–10) 2. TMS at 2 weeks Stinear, C. M. et al. Brain 2012 Aug;135:2527-35
Copyright restrictions may apply.
3. MRI/DTI at 2 weeks
Imaging based neurorehabilitation - II Predicting treatment effect – structural imaging
Damage to M1 pathway limits response to robot assisted therapy
Imaging based neurorehabilitation - II Predicting long term outcome – structural imaging
1. Database of (i) hi-res structural MRI, (ii) language scores and (iii) time since stroke 2. MRI converted to 3D image with index of degree of damage at each 2mm3 voxel
3. A machine learning approach is used to compare lesion images to others in database and similar patients identified 4. Different ‘recovery’ curves can then be estimated for different behavioural measures Months since stroke
Imaging based neurorehabilitation - II
Speech production score
Speech production score
Predicting long term outcome – structural imaging
Months since stroke
Hope et al, NeuroImage Clinical 2013
Imaging based neurorehabilitation - II Differences in residual functional architecture affected side
A
10 days post stroke
infarct
B
17 days post stroke
24 days post stroke
31 days post stroke
3 months post stroke
affected side
OUTCOMES
Barthel
ARAT
GRIP
NHPT
Patient A
20/20
57/57
98.7%
78.9%
Patient B
20/20
57/57
64.2%
14.9%
Imaging based neurorehabilitation - II Differences in residual functional architecture
unaffected
+
affected
-
unaffected
+
Predicting the effects of NIBS?
affected
-
Imaging based neurorehabilitation - II Predicting treatment effect – functional imaging
Cramer et al., Stroke 2007; 38: 2108-14
Less activity in M1 limits response to robot assisted therapy
Imaging based neurorehabilitation - II Predicting treatment effect – multimodal? •
17 chronic stroke patients (FM 4-25) 30 days of UL training – 30 mins +/- APBT
•
Change in FM relatively small 0 – 6 points
•
Overall FA (DTI) symmetry predicted ΔFM (r2 = 0.38)
•
But better model by considering those with and without MEPs
Copyright restrictions may apply.
Stinear, C. M. et al. Brain 2007 130:170-180
Imaging based neurorehabilitation - II Predicting treatment effect
Needs to inform…… ‘what kind of treatment?’ not ‘who should we treat?’
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – the key to recovery? Rehabilitation is a process of active change by which a person who has become disabled acquires the knowledge and skills needed for optimum physical, psychological and social function
Treatments aimed at reducing impairments
(Task-specific) training cortical stimulation
other
drugs
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – the key to recovery?
Drugs
NIBS
Other
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – the key to recovery?
less disability
more disability
amphetamine Several agents considered: •
Acetylcholinesterase inhibitors
•
Amphetamine
•
SSRIs (e.g. FLAME, FOCUS in UK)
•
DA agonists (e.g. DARS in UK)
Reduced GABAergic inhibition? Increased glutamatergic/BDNF mediated LTP?
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – the key to recovery?
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – the key to recovery?
Why not perform large RCTs? Inhibitory TBS?
Excitatory TBS?
Hamada M et al. Cereb. Cortex 2013;23:1593-1605
TBS (and TDCS) is very variable!
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – the key to recovery?
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – the key to recovery? Getting plasticity enhancement into clinical practice in stroke
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – the key to recovery?
ctDCS to contralesional M1 reduced SICI (less inhibition) in ipsilesional M1
tDCS-induced enhancement of skill acquisition
Reduced intracortical inhibition re-opens periods of plasticity in chronic stroke?
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – a range of scales
Intracortical networks
MESOSCOPIC
Task related networks
Large scale networks
MACROSCOPIC
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – macroscopic scale?
Network connectivity with Graph Theory for fMRI/MEG
graph metrics - efficiency
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – macroscopic scale? Dynamic Causal Modelling of motor network after stroke
Grefkes, Neuroimage (2009)
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – mesoscopic scale? “…the spectral characteristics of MEG recordings provide a marker of cortical GABAergic activity”
BASELINE BETA-BAND POWER
POST-MOVEMENT REBOUND
• Greater baseline beta-power = more inhibition?
• Greater rebound in beta power after grip = more inhibition?
• Increased by diazepam (GABAA effect?)
• Increased by tiagabine, but not diazepam (GABAB effect?)
• Increased by cTBS (decreases excitability) • Increased with ageing
MOVEMENT RELATED BETA-DECREASE
• Greater decrease in beta-power with grip = more inhibition? • Greater MRBD with diazepam and tiagabine (GABAA effect?) Rossiter et al., Neuroimage 2014
• Less MRBD in chronic stroke patients (particularly those with more impairment)
Rossiter et al., J Neurophysiol 2014
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – mesoscopic scale?
inhibitory cTBS
reduces MEPs
Biomarker
increases Beta power
Behaviour
increases RTs
Imaging based neurorehabilitation - III Enhancing Neuroplasticity – a range of scales Platform for stratification ...
Mechanistic framework ...
Predictions
macroscopic
behaviour
motor
biomarkers
language
cognitive
interventions
mesoscopic
biomarkers
patients
Bridge the gap ...
stratification
+
Imaging based neurorehabilitation Summary Early prediction of outcome
Prediction of therapy effects
Patient pathway
Prediction of plasticity modification
Imaging based neurorehabilitation Summary • Advances in neurorehabilitation are coming about through advances in neuroscience • The dose of treatment is critical - more is generally better • Enhancement of plasticity is possible
• Neuroimaging should help in stratification • Understanding the mechanisms of recovery and treatment might allow targeted or individualised therapy in future
Rehabilitation & Neuroimaging Acknowledgements FIL:
ABIU/NRU:
SOBELL DEPARTMENT :
Karl Friston
Diane Playford
Holly Rossiter
Rosalyn Moran
Fran Brander
Stephanie Bowen
Richard Frackowiak
Kate Kelly
Muddy Bhatt
Will Penny
Alan Thompson
Ella Clark
Jennie Newton
Marie-Helene Boudrias
Chang-hyun Park Sven Bestman John Rothwell Penny Talelli
Slides at www.ucl.ac.uk/ion/departments/sobell/Research/NWard
FUNDING:
