Trattografia probabilistica e fMRI nella pianificazione neurochirurgica Domenico Lizio SC Fisica Sanitaria, ASST Ospedale Niguarda, Milano domenico.lizio@ospedaleniguarda.it
Basic Anatomy The brain is full of Grey Matter neurons. These are organised into two types of “ tissues ”: - Grey Matter (GM) - White Matter (WM) Post-Mortem MRI White Matter
Basic Anatomy Neurons are densely connected and have many dendrites Axons conduct electrical signals and are surrounded by myelin Myelin is a major factor in determining the MR signal and contrast
Diffusion MRI Structural MRI Applications of MRI in Neuroimaging Mean T1-weighted diffusivity (MD ) fMRI: finger-tapp. mano dx Fractional anisotropy (FA) T2-weighted Functional MRI Principal diffusion Proton Density direction (PDD)
Applications of MRI in Neuroimaging • Study of taxi drivers showing structural plasticity Structural MRI Maguire et al.,PNAS, 2000
• Asperger Syndrome Diffusion MRI • White matter integrity - imaging tissue nature change • Differences in brain connectivity in multiple tracts Green = white matter tract ; Red/Yellow = statistically significant change in FA Roine et al., Molecular Autism, 2015
• Parkinson’s Disease Look at tracts connected to regions Structural Diffusion of structural change MRI MRI Menke et al., Brain 2013
• Stroke Therapy Single subject: responder Functional Group: MRI (Task) Correlations with improvement Johansen-Berg, et al., Brain 2002
• Altered functional connectivity Functional MRI (Resting)
Functional • Surgical planning MRI (Task) Diffusion Structural Bartsch et al., JMRI 2006 MRI MRI
Functional MRI • Measures haemodynamic response to neural activity • Task-based or resting-state-connectivity • Intrinsic contrast BOLD (Blood Oxygen Level Dependent ) • Take many fast images (EPI): 5-30 min scan
BOLD Effect • Oxy-haemoglobin Magnetic field perturbed Dephasing of nearby spins Diamagnetic Loss of signal (same as tissue) • Deoxy-haemoglobin Paramagnetic Δχ≈0.2 ppm
BOLD Effect Increased Neuronal Activity CBf Magnetic field less perturbed Less dephasing More signal CBV CMRO 2
Analysis of Functional MRI Task FMRI Resting-State fMRI & Connectivity
Images - Low Resolution FMRI • A sequence of low resolution T2*-weighted volumes are taken during the FMRI experiment • Optimised for BOLD sensitivity and speed • Take one volume every 1-3 seconds • Often take around 200 volumes (10 minutes) • An fMRI volume is shown here in orthogonal view
Functional MRI Artefacts Hardware-related Physiological Distortion Signal Loss Physiological Noise ... plus most diffusion artefacts (not eddy currents) and all structural artefacts • Distortion due to B 0 inhomogeneity (air in sinuses) - acquisition-and-analysis related fixes needed (fieldmap) • Physiological noise is more problematic near the brainstem - acquire physiological measurements & do analysis fix • Motion can also be a significant problem (some analysis fixes)
Functional MRI ‘‘Limitations’’ • Does not measure electrical activity • Does not measure metabolic activity • BOLD-FMRI is qualitative • Sensitive to fast imaging artefacts • Need good T2* sensitivity • causes lost signal in inferior regions
Functional MRI Acquisition Basic tips for acquisition • Use optimised sequences/protocol for your scanner/site • Get fieldmap (B0) for compensating distortion/signal-loss - blip-up-blip-down is not an option for functional MRI • For inferior-frontal/temporal areas apply acquisition techniques to minimise signal loss - e.g. thin slices, slice angulation, z-shims, parallel imaging, ... • Isotropic voxels (or close) are better for analysis generally • For small FOV also take one single whole-brain EPI • Biggest interaction of exp. design-acquisition-analysis so think carefully about all parts before acquiring data!
Diffusion MRI • Measures microstructure directionality and “ integrity ”, particularly in WM • Provides information on anatomical connectivity • Need to acquire many “directions”: 5 -30 min scan
What is diffusion? • Random motion of particles due to thermal energy • Water molecules collide and experience net displacement • Displacement described by diffusion coefficient (D) • Normally, diffusion is isotropic (equal in all directions)
Why is diffusion interesting? • Diffusion is restricted by tissue boundaries, membranes, etc • Marker for tissue microstructure (healthy and pathology)
Diffusion anisotropy in white matter Water can diffuse more freely along white matter fibres than across them
Diffusion anisotropy in white matter Diffusion in white matter fibres is “anisotropic” Directionality of diffusion tells us about fibre integrity/structure and orientation
The diffusion tensor Displacement due to diffusion is approximately ellipsoidal Eigenvectors = axes of ellipsoid (direction of fibres) Eigenvalues = size of axes (strength of diffusion)
The diffusion tensor: Useful quantities Principal diffusion direction (PDD): what direction is greatest diffusion along? Info about fibre orientation Fractional anisotropy (FA): how elongated is the ellipsoid? Info about fibre integrity The diffusion tensor: Useful quantities Mean diffusivity (MD) Info about tissue integrity
Diffusion tensor imaging Mean Fractional Principal diffusivity (MD ) anisotropy (FA) diffusion direction (PDD) At each voxel, fit the diffusion tensor model Can then calculate MD, FA, PDD from fitted parameters
Mean diffusivity (MD) Control MD Acute Stroke Mean diffusion coefficient across all directions Correlate of tissue integrity (white and gray matter) Example: MD is altered in acute and chronic stroke
Fractional Anisotropy (FA) Inequality of diffusion coefficient across different directions High in regions where diffusion is most directional Relates to integrity of white matter fibre bundles
Principal diffusion direction (PDD) Direction along which greatest diffusion occurs Relates to direction of fibre orientations Typically, will use this as starting point for fibre tracking
Diffusion tractography Follow PDD to trace white matter fibers (“tractography”)
More complicated models: Crossing fiber populations
Diffusion tractography: Determionistic vs Probabilistic Deterministic assumes a single orientation at each voxel: one • streamline per seed voxel. Probabilistic assumes a distribution of orientations: multiple streamline • samples per seed voxel (drawn from probability distribution)
Acquiring the image If diffusion is present, gradients cause a drop in signal. Greater Diffusion = Less Signal
Diffusion MRI Artefacts Hardware-related Eddy Currents + Bulk/Pulsatile Motion … plus all the structural artefacts • Eddy currents: both acquisition and analysis fixes available • Distortion due to B 0 inhomogeneity (air in sinuses)- acquisition- and-analysis related fixes needed • Bulk motion is corrected for in acquisition (navigators) • Pulsatile motion is more problematic
Motion in Diffusion MRI Linescan diffusion image Diffusion gradients encode tiny displacement Subject motion is also accidentally encoded Image artefacts if we try to combine data from multiple excitations (different motion)
Can motion be avoided? Subject restraints can reduce bulk motion, but... ...in the brain, there is significant non-rigid motion from cardiac pulsatility cardiac gating helps, but brain is never very still!
Single-shot echo-planar imaging (EPI) Single-shot imaging freezes motion Most common method is echo-planar imaging (EPI) Images have serious distortion and limited resolution
Eddy Currents Eddy currents “resist” gradient field changes effective gradient fields Diffusion gradients create large eddy currents, which persist into acquisition window Distort the k-space trajectory, casing shears/scaling of images
Diffusion MRI “Limitations” • Does not measure axon size/density directly • Does not measure single fibres (only average groups) • More difficult to deal with crossing/kissing fibres • Quantitative local measurements, but not connectivity • More difficult to do in pulsatile regions (e.g. brainstem) • More restricted by hardware and SNR • Sensitive to fast imaging artefacts
Diffusion MRI: Acquisition Basic tips for acquisition Best parameters can be quite hardware dependent (esp. • gradients) so check what is optimised for your scanner In general, b-value of 1000-1500 s/mm2 and 60+ directions • (tractography) or 12+ directions (FA, etc.) tend to give good results (but the more directions the better) Get one b=0 image for every 8-10 diffusion-weighted images • Get a fieldmap (B0) for distortion correction - or alternatively, a • blip-up-blip-down • Isotropic voxels (or close) are better for analysis generally Do not do oversampling on scanner (sometimes the default) • Both single/multi-shell give good tractography •
Recommend
More recommend
