MR Diffusion, DSI, DTI and Tractography
The MR Diffusion Imaging (DTI), Diffusion Spectrum Imaging (DSI), Diffusion Tensor Imaging (DTI) and Tractography web based training will focus on the applications, sequences, and parameters as well as defining these various techniques.
Define Diffusion and identify the differences between Isotropic and Anisotropic diffusion Identify the diffusion sequence and key features of the diffusion parameter cards Explain Diffusion RESOLVE Explain Simultaneous Multi-Slice (SMS) EPI Explain QuietX Diffusion (qDWI) Explain Diffusion Spectrum Imaging (DSI) Define Tensor Explain Tractography Diffusion, Diffusion Spectrum Imaging, Diffusion Tensor Imaging and Tractography Form of MR imaging based upon measuring the random Brownian motion of water molecules within a voxel of tissue.1 Also known as Brownian motion (Robert Brown - botanist) Robert Brown2 - Botanist 1 http://radiopaedia.org/articles/diffusion-weighted-imaging-1; 2 http://www.famousscientists.org/robert-brown/; MAGNETOM Maestro Class Diffusion Weighted MRI of the Brain – Siemens Medical Isotropic Uniform in all directions (e.g., sphere) Molecules move randomly without restrictions (e.g., molecule in a glass of pure water) Anisotropic Directionally independent (e.g., ellipsoid) Diffusion in orientated tissue Molecules move freely in the brain, but are inhibited by structures (e.g., cell membranes, fiber tracts, myelin sheaths, etc.) Routine Applications Early detection and assessment of stroke Differentiate new from old infarcts Tumor characterization Evaluation of Multiple Sclerosis Advanced Applications Tractography Tensor Echo Planar Imaging (EPI) Single RF Excitation pulse All k-space filled using multiple gradient echoes created by an oscillating frequency-encoding gradient Each echo phase encoded separately using a low amplitude phase encoding gradient (blip) The number of k-space lines collected in a single shot is referred to as the E-P-I factor. Employs gradients with equal amplitude, high enough to be sensitive to microscopic motion, but opposite polarity Duration and Amplitude of gradients defines b-value Requires only 1 measurement per slice position and b-value Shorter acquisition Diffusion vectors [x, y, z] Trace-weighted images generated in 3 diagonal axes and averaged Shorter TE and higher SNR than 1-scan Trace 1 diffusion weighted image per slice position and b-value is calculated 4 Measurements per image Shorter TE Higher SNR than 3-Scan Trace TE = 51ms, iPAT 3 3 scan trace, 3 averages TA = 59s TE = 51ms, iPAT 3 4 scan trace, 3 averages TA = 1 1min, 16s SNR and Scan Time comparison 3 Scan Trace = 3 Averages 4 Scan Trace = 3 Averages Acquires 3 orthogonal images per slice position and b-value (b=0) 1 image for each for diffusion direction (read, slice and phase) b=0 – 1 image per slice direction Diffusion vectors – read, slice, and phase Detects movement in only one direction Acquires 1 image per slice position and b-value No trace-weighted images Diffusion vectors – phase, read, slice Slice – Diffusion Weighting in the Slice direction Read – Diffusion Weighting in the Readout direction Phase – Diffusion Weighting in the Phase direction Acquires 1 diffusion-weighted image per slice position, per b-value, and (for b>0) per diffusion encoding direction Defines number of directions with the Diffusion Directions parameter per slice position and b-value Diffusion vectors [x, y, z] Acquires 1 image per slice position and b-value Diffusion images with minimum TE Diffusion weighting in the direction of the spatial diagonal No trace-weighted images Diffusion vector – [x, y, z] Bipolar Double refocused spin echo diffusion encoding Minimizes distortions MR signals of different contrasts are alternately read out with a positive or negative gradient Monopolar Single refocused spin echo diffusion encoding Allows for shorter T-E with an increase in signal-to-noise MR signals of all contrasts are read out with a positive gradient Bipolar Double refocused Spin Echo diffusion encoding Minimizes distortions Monopolar Single refocused Spin Echo diffusion encoding Allows ↓TE → ↑SNR Diffusion Weightings Defines number of diffusion-weightings acquired during a measurement Defines individual weightings in b-value input field prior to the measurement e.g., Maximum number of b-values = 16 b-value Measure of diffusion-weighting Units – s/mm2 The greater the b-value the stronger the diffusion weighting Increases with intensity, duration, and time interval of the diffusion-sensitive gradient pulses b-value = 0 – corresponds to one T2*-weighted image Defines number of averages per b-values Protocols with multiple b-values – individual number of averages may be selected for scans with different diffusion weightings (b-values) Used to optimize acquisition time DWI – less averages for b=0; more averages for b > 0 DTI – more averages for b=0; less averages for b > 0 Diffusion weighted Images Determines if original images with diffusion weighting are reconstructed Images are a combination of T1, T2, and diffusion-weighting Trace weighted images Isotropic diffusion-weighted image is reconstructed Diffusion-weighting is applied in all 3 spatial directions e.g., Parameter availability depends on Diffusion Mode selected Calculated when averaging is performed with different b-values Show the diffusion coefficient as grayscale values Free of T1 and T2 contributions Displays pure diffusion contrast e.g., eliminates t2 shine-through Minimum of two b-values required Threshold value to determine the pixel intensity from which ADC maps are calculated Pixel values with an intensity value lower than the “Noise Level” entered will not be taken into account Calculated as the ratio between the signal after applying diffusion gradients and the signal without the diffusion gradients is applied Similar to Trace Weighted but it eliminates T2 shine-through artifacts Available only for syngo Classic Not available with Diffusion mode MDDW Inverted gray Scale Inverted (PET-like) display of the diffusion trace weighted image Calculated Image Determines whether a virtual (b-value) image is calculated b-value for calculation is set in Calculated bValue Specifies minimum b-value used in ADC calculations Function may be useful for diffusion-weighted imaging of the abdomen Supports calculation of images that are insensitive of vascular capillary perfusion Reduces eddy current induced distortions of diffusion weighted images Only available for Single-Shot EPI Diffusion Direct Mode – Registration of individual diffusion encoded images to undistorted reference image (Image b=0 s/mm2) Additional b0 images acquired independent of protocol settings Recommended for diffusion weighted images with sufficient SNR Adjustment Mode – Dedicated, additional adjustment scans to estimate dynamic field errors Recommended for images with low SNR images (high b-values) Direct Mode Reference (b=0) Distorted Image Corrected Image Fat Suppression Pulse – ON (e.g., required) Fat Suppression – OFF Causes chemical shift effects due to: Insufficient fat suppression Low bandwidth phase direction High Bandwith Low Bandwidth Diffusion Sequence Fat Suppression Fat Suppression eliminates chemical shift artifacts in phase encoding direction Fat Sat mode – “Strong” recommended Improved fat suppression using an inverted slice selective gradient to reduce ghosting artifacts due to imperfect fat suppression Phase Encoding: A to P Phase Encoding: R to L Default Anterior to Posterior Reduces artifacts in phase encoding direction Keeps image symmetrical Reduces nerve stimulation Phase Encoding: A to P SP F13.2 Phase Encoding changed: P to A SP F13.2 Diffusion Phase Encoding Direction Changed to Posterior to Anterior Reduces geometric distortions depending on the slice position Moves artifact ouf of area of interest Phase Encoding: A to P SP F0.2 Phase Encoding changed: P to A SP F0.2 ADC Inline – Automated post-processing ADC Offline – Manual Post-processing Workflow Open Patient Browser Select diffusion sequence Evaluation > Dynamic Analysis > ADC ADC window opens ADC Offline – Manual Post-processing Workflow - ADC Window Noise Level – 40, typically used for Brain exams Select > OK – Starts post-processing Images stored to Patient Browser ADC window features Test Image – Optional Select pull-down and center image number Select Test (e.g., provides preview ADC image) Noise Level – threshold value determines pixels used for evaluation (e.g., defaults to noise level number entered in the diffusion parameter card) Result Series Description – default series name for Patient Browser syngo RESOLVE - Readout Segmentation of Long Variable Echo-trains Readout-segmented (multi-shot) diffusion-weighted EPI High-resolution DWI and DTI Including inline maps and post-processing with Neuro 3D (MR) Task Card Reduced susceptibility and blurring artifacts due to reduced TE and Echo Spacing Insensitivity to motion-induced phase errors Reduced SAR in comparison to TSE-based methods 2D navigator phase correction with automatic re-acquisition of data with large phase errors Diffusion 3 Scan Trace RESOLVE 3 Scan Trace Resolution Parameter Card Readout segments – Number of segments in the readout direction for multi-shot, readout-segmented sequences Sequence Parameter Card > Part 2 Tab Reacquisition mode – Defines if acquisition is automatically repeated to improve image quality when data is corrupted (e.g., motion) ON – recommended for Neuro syngo RESOLVE studies syngo RESOLVE (VE11) Partial Fourier in Readout Direction Reduces number of shots required for each image Shorter scan times Potential signal-to-noise loss RESOLVE Readout Partial Fourier: 7/8 4/5 RO segments acquired, 27 slices Scan time: 2:03 mins ss-EPI 27 slices Scan time: 1:38 mins. RESOLVE – Partial Fourier in Readout Direction Tool tip – float mouse over Readout Partial Fourier displays number of acquired readout segments for each setting Note – some cases, different settings result in the same number of acquired settings Adaptive Coil Combine Mode – Set to Adaptive Combine. System > Miscellaneous parameter card Default, Siemens RESOLVE programs Reference: Siemens RESOLVE Hot Topic A911M-MR-141700-P1-4A11.pdf Synonyms SMS = Multi-Band (MB) Slice Acceleration Factor = SMS = MB Factor Functionality Excitation of multiple slices simultaneously Separation of slices with different coil sensitivities along the slice direction (similar to iPAT) Benefits Faster imaging More slices Shorter TR Improved Temporal Resolution Higher Spatial Resolution Blipped CAIPIRINHA Creates Field-of-View (FoV) shifts of parallel slices in phase encoding direction Minimizes g-factor related to Signal-to-Noise Ratio (SNR) losses during reconstruction of individual imaging slices Sequence options SMS EPI DWI SMS EPI BOLD Note: Optional sequence MAGNETOM Prisma Head/Neck 64 coil MAGNETOM Skyra Head/Neck 20 coil SMS EPI DWI Reduces imaging time High resolution diffusion tensor imaging Routine Diffusion-weighted imaging Benefits Higher spatial resolution Higher diffusion resolution Increased diffusion directions Increased b-values Resolution > iPAT Parameter Card Accel mode Accel factor slice Special readout segmented EPI sequence Stretched gradients to minimize the slew rate Reduces acoustic noise during scanning Provides image quality comparable to single shot EPI Sequence > Part 2 Parameter Card Acoustic noise reduction Always active – triggers quiet adjustments Resolution > Common Parameter Card Phase Partial Fourier – selectable for shorter TE Sequence location Dot Cockpit > Head > Programs > Library > Quiet > qDWI License required – Advanced Diffusion qDWI EPI Advanced diffusion technique to resolve fine anatomical details of the brain Crossing white-matter fibers by using multiple diffusion directions and b-values in a single measurement Diffusion Mode – q-Space Enables acquisition of corresponding data Employs Cartesian sampling pattern Permits isotropic sampling of a spherical region in q-Space with user defined step size Planning q-Space measurement Select Diffusion sequence (ep2d_diff) > Diffusion Neuro Parameter Card > Diffusion Mode – q-Space q-Space specific measurement parameters q-Space weightings – number of q-space coordinates sampled along each positive coordinate axis (including the origin) q-Space maximum b-value – b-value that corresponds to the diffusion weighting of the outermost q-space coordinates q-Space coverage Full – Complete q-space coverage (Sampling of a spherical q-space region) Half – Partial q-space coverage (only ½ of the sphere is sampled, reducing acquisition time) What is a Tensor? Tensor is a complex mathematical equation used to model 3D anisotropy Diffusion Tensor Imaging (DTI) Provides a 3 dimensional description of diffusion Requires calculation of a Tensor Displays direction and magnitude of anisotropic diffusion Preferential diffusion direction infers white matter tract structure Diffusion in 3 Orthogonal Planes (X, Y and Z axis) Diffusion Tensor – information on multiple directions (e.g., Diffusion directions – minimum = 6) Diffusion Ellipsoid Diffusion D in Acquisition Coordinates e1,e2,e3 Eigen Values Co-ordinate System Reference: 1http://psychcentral.com/lib/what-is-functional-magNnetic-resonance-imaging-fmri; 2 http://medical-dictionary.thefreedictionary.com/blood+oxygen+level+dependent MDDW must be selected to enable Diffusion Directions Diffusion Directions Minimum number of Diffusion Directions for MDDW = 6 Number of Diffusion-encoding directions = 6, 10, 12, 20, 30, 64, 256 Trace Weighting Geometric averaging of all DTI directions Available for both (syngo Classic and syngo.via) Apparent Diffusion Coefficient (ADC) maps Show the diffusion coefficient in grayscale values Free of T1 and T2 contributions Display pure diffusion contrast Eliminate T2 shine through Available for both syngo Classic and syngo.via FA Maps – (Fractional Anisotropy Maps) Determines if an FA map is reconstructed Isotropic diffusion characteristics are shown as dark Anisotropic diffusion characteristics are shown as bright FA Color maps Show direction of anisotropy syngo classic - Neuro 3D MR Task Card - Available inline or offline syngo.via - Available offline FA Maps – (Fractional Anisotropy Maps) Calculates one image from a random number of orientations Spatial information, minimum of 6 gradient directions, preferably 12 or more have to be acquired Tensor glyphs – Spatial orientation of preferred direction Sphere = Isotropic diffusion Ellipse or Cigar shape = Anisotropic diffusion Available for both FA Gray scale and FA Color Map (syngo Classic and syngo.via) FA Maps – (Fractional Anisotropy Maps) FA Color Map – Directions along main axis are color-coded Red = L>>R Medio-Lateral Blue = H>>F, Cranio-Caudal Green = A>>P, Anterior-Posterior Mosaic One Mosaic image per measurement One series for all images Tensor Diffusion Tensor data file is reconstructed and stored on the Patient Browser Post-process DTI (Tensor) with syngo.via or syngo Classic (Neuro 3D (MR) Task Card Control Mode for DTI B1 control loop implemented for DTI – ensures high signal stability throughout the measurement Reference amplitude continuously measured and updated Transmit reference amplitude adjusted for a highly stable signal intensity Available for ep2d_diff sequences Active for DTI measurements Definition Diffusion Weighted sequence (MDDW) 3D Modeling Technique Visually represent neural tracts Post-processing syngo.via (or) syngo Classic Neuro 3D MR Task Card Define Diffusion and identify the differences between Isotropic and Anisotropic diffusion Identify the diffusion sequence and key features of the diffusion parameter cards Explain Diffusion RESOLVE Explain Simultaneous Multi-Slice EPI Explain QuietX Diffusion Explain Diffusion Spectrum Imaging Define Tensor Explain Tractography
- MR Diffusion
- DSI
- DTI and Tractography