3.1 Physiological MRI Scan Session

This session includes 38 minutes of image acquisition and should be completed in less than one hour of total scan time. This session includes survey scan for head position and positioning image planes subsequent in subsequent scans, a reference scan for calibration of the 32 channel SENSE head-coil, a 3D T1 weighted (1,2 or IR Prepped) high-resolution anatomical scan of the entire brain as well as 3D T2 weighted anatomical scan (FLAIR) of the same brain region, also with spatial resolution of 1 x 1 x 1 mm. In addition, quantitative perfusion image data is acquired in the resting state using a pCASL method. In order to accurately quantitate the baseline perfusion in each subject, a series of T1 and transit time measurements are acquired along with a measurement of the ASL labeling efficiency. The sequence of scans is tabulated directly below. Details of the key imaging parameters for each sequence are listed in the text that follows. The Table below provides numerical values for key image parameters for each sequence.

List of Scans in the Physiological Session (Total Time: 38:18.8)
Scan name Scan time             Scan name Scan time
Survey 00:31   T1EST_TI400 00:20.0
Ref_Head_32 00:44.4   T1EST_TI1000 00:20.0
T1W_3D_IRCstandard32 05:15.5   T1EST_TI100 00:20.0
EPI_BOLD_NR1 00:10.0   T1EST_TI3000 00:20.0
ALPHA MEASUREMENT 04:48.0   T1EST_TI300 00:20.0
Baseline CBF 04:21.8   T1EST_TI1500 00:20.0
T1EST_TI200 00:20.0   T1EST_TI500 00:20.0
T1EST_TI750 00:20.0   T2W_3D_FLAIR_1x1x1 SENSE 08:33.6
T1EST_TI2000 00:20.0   DTI_68_iso_32chSHC 09:53.0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Detailed Descriptions of Physiological Scans

Survey

Fast-field echo sequence – Any fast gradient echo method will do for this purpose. This scan is not used for any quantitative analysis.

The survey scan is used to assess head position in the scanner and to readjust if necessary to obtain optimal alignment of the interhemispheric fissure with the y-z plane of the scanner and placement of the thalamus near the isocenter of magnet, at the center of the image field of view (FOV). The survey scans are also used for positioning of the imaging planes for all subsequent scans in the physiologic imaging session. We use a fast gradient-echo method with a total scan time of approximately 30 seconds. Three slices are acquired in each of three orthogonal scan planes so that subsequent scanning prescriptions can be made and confirmed in any of these three planes.

 

Resolution:

Acquistion Matrix: 1x1x10
Reconstucted Matrix: 1x1x10
Acquistion Plane: 3 planes
Total scan time: 32 secs

 

 

 

 

 

 

Contrast:

TR/TE: 11/4.6 msec
NSA: 1
TFE shots: 2
TFE Shot interval: 1161 msec

 

 

 

 

T1W-3D

Turbo-field echo sequence – MPRAGE or Inversion Recovery Prepared Fast Field Echo.

 

This is a standard Inversion Recovery prepared, T1-weighted, 3D isotropic method is used to acquire a high resolution anatomical image for underlayment of functional and DTI maps. This type of scan is referred to as MPRAGE1-5 on Siemens scanners or more generically as MDEFT.6,7 The Inversion Recovery time is optimized for GM/WM contrast at TI=939 msec. This data set can also be used for brain tissue automated segmentation in SPM and other applications. Segmentation accuracy can be improved by using the T2W-3D FLAIR images in conjunction with this data set in a multi-parameter segmentation approach available in SPM or our own method. Images are acquired in the sagittal plane at a spatial resolution of 1 x 1 x 1 mm. Using a SENSE factor of 2 the total image acquisition time is 6’20”. Note that the turbo direction is set for the “Y” direction, corresponding to the A-P axis of the subject. The TFE factor of 224 is the same as the resolution in the AP direction.8

 

Key Imaging Parameters:

Resolution: 1x1x1
Acquistion Matrix:

1x1x1

Reconstucted Matrix: 1x1x1
Acquistion Plane:

Sagittal

Total scan time:

5 min 15 secs

 

 

 

 

 

 

 

Geometry:

FOV: (default but can vary for each subject.)
  FH(mm): 256
  AP (mm): 224
  RL (mm): 160
  Shim: PB Volume (i.e. highest order shims possible within one volume)
SENSE: Yes
  P-Reduction (AP): 1
  Pos Factor: 1
  S-Reduction(RL): 2
 Phase Encoding Direction: AP

 

 

 

 

 

 

 

 

 

 

 

Contrast:

FA:
Shot interval: 2800 ms
TI: 939 ms
TR/TE: 8.1 / 3.7 msec
TFE shot duration / Acq: 1851.9/1804.6 ms
NSA: 1

 

 

 

 

 

 

 

EPI-BOLD_NR1

Gradient-echo, single shot EPI acquisition.

 

This is a typical T2* weighted EPI gradient echo scan used for BOLD-fMRI. In this case we acquire a single time-point with 24 axial slices covering the entire brain. The primary purpose of this scan is to cause the higher order shimming to occur in the selected brain volume and to define this functional volume for use in subsequent EPI based scans for optimal field uniformity. This is a fast scan with the same resolution and field of view that will be used in the baseline CBF measurements. Inspection of this image will be done by the operator to determine whether sufficient homogeneity has been achieved by the shimming process in order to proceed to obtain high quality images. If these images demonstrate distortion or other artifacts, shimming should be repeated or other measures taken to improve the patient positioning and image quality. High order shimming performed to optimize this scan volume should be saved and applied to all subsequent scans in the Physiological Sequence. On the Philips platform this is done by propagating the PB-Volume shim parameter from one scan to the next and preserving the scanning geometry.

 

Key Imaging Parameters:

Bandwidth EPI:

3313.5 Hz

Acquistion Matrix: 64x64
Total scan time: 10 secs

 

 

 

 

Geometry:

FOV: (default but can vary for each subject.)
  RL (mm): 240
  AP (mm): 240
  FH (mm): 120
SENSE: Yes
  P-Reduction (AP): 2
Phase Encoding: AP
Reconstructed Matrix: 64x64
Resolution: 3.75x3.75x5
Stacks: 1
Slices: 24
Slice Orientation: Transverse

 

 

 

 

 

 

 

 

 

 

 

 

Contrast:

FA: 90°
TR/TE: 2000 / 35 msec
NSA: 1
Fat Suppression: SPIR
Shim: PB Volume (i.e. highest order shims possible within one vol)

 

 

 

 

 

 

Labeling Efficiency (Alpha) Measurement

Multi Shot EPI Gradient Echo Measurement

This is a single-shot gradient echo EPI image acquired in a single slice, which is placed above the carotid bifurcation . The results of this image enter into the formula for quantitative CBF estimation from the ASL data, so an accurate assessment of labeling efficiency is important. For this test, the label slab should be placed transversely just below the carotid bifurcation with the acquisition slice 30 mm above this position, placing it in the carotids before they connect with circle of Willis. See diagrams in procedure manual for more details on label slab placement.

The mean B1 amplitude to 1 mT results in a pCASL flip angle of approximately 22o.9 Each pulse in the pCASL pulse train has a duration of 500 msec and a repeat interval of 1.5 msec. This leaves a gap between pulses of 1.0 msec. With the 22oflip we do not run into the SAR limit. The mean PCASL Labeling gradient strength is optimized for label saturation with a value is = 1.00 mT/m. These values are now used consistently for all ASL methods in the protocol. This produces maximum label efficiency and also checks out with Luis Hernandez-Garcia computer simulation.10

Labeling efficiency is computed from these data by using the magnitude and phase images from each label delay value to form a complex image. Control image phase is used as reference and all images referenced to it. Label and control scans are subtracted (complex). Real part of this difference is used as an estimate of the labeling efficiency (alpha)

Alpha = |C-T| / 2|C| : where T is the tagged image and C is the control image. Note that these are complex images (indicated by the BOLD fonts) so complex subtraction must be used in the numerator. This is implemented in the ASL perfusion processing module for the LONI Pipeline.

 

 

Maximum difference is selected in the single slice acquisition in the center of the carotid artery. This single voxel point is used for the calculation of Alpha (a). Alpha (a) is used in the magic formula for calculation of CBF.11,12 Note, another approach would be to use an ROI in the carotid or to use voxels from right and left carotid arteries averaged together to obtain a more robust estimate of a.

Here Mc is the average control image intensity, Ml is the average label image intensity, t is the duration of the labeling pulse, and w is the post-labeling delay time, which will be adjusted on a slice-by-slice basis (if slice timing correction was not previously employed) to account for the slice timing of the EPI acquisition.

The other parameters are physical constants:

T1a = T1 of blood, l is the blood-tissue water partition coefficient, a is the tagging efficiency, d is the tissue transit time, T1is the T1 of tissue, and T1rf is the T1 of tissue in presence of RF. The defaults will be the published values in the literature 13 of R1a ( = 1/T1a) = 0.67 s-1, a = 0.68, l = 0.9 g/ml; however, the user will have the option of interactively changing them in the processing pipeline if more accurate, measured values are available. 

 

Key Imaging Parameters:

Acquistion Matrix: 132x126
EPI Bandwidth: 1211.4 Hz
Total scan time: 4 min 48 secs

 

 

 

 

Geometry:

FOV: (default but can vary for each subject.)
  RL (mm): 200
  AP (mm): 200
  FH (mm): 3
SENSE: No
Phase Encoding Direction: AP
Reconstucted Matrix: 144x144
Reconstructed Voxel: 1.39x1.39x3
Acquistion Voxel: 1.52x1.59x3.0
No. of Shots: 14
No. of echoes per shot: 9
No. of dynamic scans: 10
No. of slices: 1
No. of Packs: 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Contrast:

FA: 90°
TR/TE: 1000 / 6.9 msec
NSA: 1
Shim: PB Volume (i.e. highest order shims possible within one vol)

 

 

 

 

 

ASL Parameters:

ASL: pCASL
Label Type: Parallel Slab
Label Gap: 5mm
Label Duration: 500 ms
Post Label Delay: 5 ms
FLL pCASL B1 amplitude: 1.0 uT
FLL pCASL interval: 1.5 ms
FLL pCASL flip angle duration: 0.5 ms
FLL pCASL gradient strength: 6 mT/m
FLL pCASL mean gradient: 1 mT/m

 

 

 

 

 

 

 

 

 

 

T1Mapping – T1

Inversion Echo Single-Shot EPI Gradient Echo

A T1 estimate is also needed in the formula described above for estimating CBF.11,12  It is important to estimate T1 on a pixel by pixel basis to get accurate estimates of CBF from the model because T1 varies over the brain and between GM and WM. T1 also varies with age, so it is important to map it in each subject. The T1 variation has a significant effect CBF estimate from the Magic Formula (>20% errors) so it must be measured voxel wise in each subject in order to get accurate CBF values. We have implemented a fast T1 measurement method using an inversion recovery EPI gradient echo sequence. A series of 10 separate T1-weighted EPI scans with inversion times as follows: 100, 200, 300, 400, 500, 750, 1000, 1500, 2000, 3000 msec. These scans are done using the same FOV and resolution that is used for the baseline CBF measurements so that T1 estimates can be computed on a voxel-wise basis for each voxel in the CBF maps. 

 

Key Imaging Parameters:

Acquisition Matrix: 64x64
Bandwidth: 2944.5 Hz
Total scan time: 20 secs

 

 

 

 

Geometry:

FOV: (default but can vary for each subject.)
  FH(mm): 120
  AP (mm): 240
  RL (mm): 240
SENSE: Yes
  P-Reduction (AP): 2
  Pos Factor: 1
Phase Encoding: AP
Stacks: 1
Slices: 24
Reconstructed Matrix: 64x64
Resolution: 3.75x3.75x5

 

 

 

 

 

 

 

 

 

 

 

 

Contrast:

TR/TE: 5000/23 msec
IR delays (msec): 100,200,300,400,500,750,1000,1500,2000,3000
NSA: 1
Fat Suppression: SPIR
Shim: PB Volume (i.e. highest order shims possible within one vol)

 

 

 

 

 

 

Baseline – CBF (ASL)

PCASL tagged EPI single shot gradient-echo acquisition.

A key deliverable element of the Pediatric Functional Neuroimaging Research Network is data documenting the cerebral perfusion in the developing brain in children ranging in age from birth to 18. This is done non-invasively, without the use of exogenous MR contrast by using a classic pCASL spin labeling pulse MRI method. 9,10,14 Based on literature values and computer simulations as well as the practicalities of acquiring this sort of data in a reasonable amount of time over the entire brain, we have established the following protocol for data acquisition across the age span. 

Arterial spin tagging is achieved using a pCASL labeling method as described above under the Labeling Efficiency method.11,15 A mean B1 amplitude of 1 mT is used resulting in a pCASL flip angle of approximately 22o per pulse as described above. Each pulse in the pCASL pulse train has a duration of 500 msec and a repeat interval of 1.5 msec. This leaves a gap between pulses of 1.0 msec. With the 22o flip we do not run into the SAR limit. The mean pCASL Labeling gradient strength is optimized for label saturation with a value is = 1.00 mT/m. These values are now used consistently for all ASL methods in the protocol. This produces maximum label efficiency and also checks out with Luis computer simulation. Labeling efficiency is estimated separately from the Alpha measurement acquisitions above.  The label duration for the baseline CBF measurement is 2000 msec and the post-label inflow delay is 1400 msec. The details are enumerated in the table below as well as in the accompanying exam card for the Philips scanner.

Following the tagging sequence, a single shot echo planar gradient echo image is acquired with the resolution and timing below. 30 dynamic scans are acquired with a control and labeling scan in each dynamic interval (TR=4000). CBF is calculated from the difference between the tagged and control images using the Magic Formula above. This difference enters the equation at DMThe control label pulses are used to control for magnetization transfer effects.16 The pulse train is identical between label and control however the phase of the pulses during the control sequence is alternating by 180o, resulting in no net precession of the spins as they traverse the inversion plane defined by the label gradient and offset frequency.

 

Key Imaging Parameters:

Bandwidth EPI:

3080.7 Hz

Acquistion Matrix: 64x64
Total scan time: 4 min 21 secs

 

 

 

 

 

Geometry:

FOV: (default but can vary for each subject.)
  RL (mm): 240
  AP (mm): 240
  FH (mm): 120
SENSE: Yes
  P-Reduction (AP): 2
  Pos Factor: 1
Phase Encoding: AP
Reconstucted Matrix: 64x64
Resolution: 3.75x3.75x5
Stacks: 1
Slices: 24
Slice Order: Ascend
Slice Orientation: Transverse

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Contrast:

FA: 90°
TR/TE: 4222 / 11 msec
NSA: 1
Fat Suppression: SPIR
Shim: PB Volume (i.e. highest order shims possible within one vol)

 

 

 

 

 

 

ASL Parameters:

No. of Dynamic Scans: 30
ASL: pCASL
Label Type: Parallel Slab
Save/Restore Labeling: Restore
Label Location: F
Label Gap: 60
Label Duration: 2000
Post Label Delay: 1400 msec
Vascular Crushing: No
FLL pCASL B1 amplitude: 1.0 uT
FLL pCASL interval: 1.5 ms
FLL pCASL flip angle duration:

0.5 ms

aFLL pCASL gradient strength: 6
FLL pCASL mean gradient: 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

T2-Weighted 3D FLAIR

This is a 3D fluid-attenuated, inversion-recovery sequence with long inversion time to suppress CSF signal and long TE to emphasize T2 contrast between GM and WM.17,18 Following the inversion pulse, images are recorded with a multi-shot turbo spin echo pulse sequence. Images are acquired using the same resolution and geometry as the 3D Tso that the data from the two contrasts can be used jointly to improve automated segmentation of GM and WM for voxel-based morphometry and normalization. Fluid inversion recovery is used to suppress CSF signal and obtain good GM/WM contrast with high dynamic range. Inversion time selected is TI=1650. K-space is optimized by the Philips software to achieve and effective echo time of TE=125 msec.

 

Key Imaging Parameters:

Resolution: 1x1x1
Acquistion Matrix: 256x224x160
Reconstucted Matrix: 1x1x1
Acquistion Plane: Sagittal
Total scan time: 8 min 34 secs

 

 

 

 

 

 

Geometry:

FOV: (default but can vary for each subject.)
  FH (mm): 256
  AP (mm): 224
  RL (mm): 160
Shim: PB Volume (i.e. highest order shims possible within one volume)
SENSE: Yes
  P-Reduction (AP): 1.5
  Pos Factor: 1
  S-Reduction (RL): 2
Phase Encoding Direction: AP

 

 

 

 

 

 

 

 

 

 

 

Contrast:

FA:
Shot interval: 2800 ms
TI: 1650 ms
TR/TE: 4800 / 293msec
TE equiv: 125 ms
TFE factor: 182
NSA: 2

 

 

 

 

 

 

 

DTI-61 Direction

Spin echo EPI single-shot sequence

In this case we use a spin-echo, EPI method with IVIM gradients for diffusion weighting of the scans. Some discussion about spatial resolution vs. angular resolution vs. b-values took place among several of the investigators and consultants in converging to the final parameters for this scan. The proponents of true HARDI scanning advocated for the use of a higher b value for more extreme diffusion weighting (b=3000) in order to allow for probabilistic fiber tractography to resolve crossing fibers.19-22 On the other hand, there are arguments favoring a larger number of diffusion directions in order to be able to resolve crossing fibers from DTI data.23 However, competing against higher b-values and higher angular resolution are the demands of SNR and acquisition time.24 From the literature and discussion with some of those actively using DTI to study early brain development in children we discovered that there are strong arguments in favor of the lowest possible b-value (700-800) for diffusion weighting combined with the highest possible spatial resolution in the EPI data. 25-28 For this study the DTI is an extra item not required by the contract as a deliverable. To avoid putting the core data elements required for the study at risk, we felt compelled not to allocate too much time to this data acquisition and we aimed for the goal of acquiring the highest possible resolution data with the best SNR possible in less than 10 minutes. This resulted in an MRI acquisition that was dervired from those distributed by Philips with the 32 channel head coil that makes use of a SENSE factor of 3. In this case, we were able to obtain 2 x 2 x 2 mm spatial resolution with 61 diffusion directions at b=1000 in under 10 minutes. This method was felt to be a good compromise following discussions with Tim Roberts, Paul Thompson, Sususu Mori, Weihong Yuan, and Vince Schmithorst. Seven b=0 images are also acquired at intervals of 8 images apart in the diffusion direction vector. These b0 images are used for corregistration and averaged to form the baseline for solving the diffusion tensor for the eigen values used in quantitative diffusion parameter calculations.

 

Key Imaging Parameters:

EPI Factor: 39 
EPI Bandwidth: 1752.6 Hz
Acquistion Voxel: 2x2.05x2
Reconstucted Voxel: 2x2x2
Acquisition Matrix: 112x109
No. of directions: 61 plus 7 b0 images = 68
Total scan time: 9 min 53 secs

 

 

 

 

 

 

 

Geometry:

FOV: (default but can vary for each subject.)
  RL (mm):  224 
  AP (mm):  224 
  FH (mm):  120 
SENSE:  Yes 
  P-Reduction (AP):  3 
  Pos Factor:  1 
Phase Encoding:  AP 
Resolution: 2x2x2
Stacks: 1
Slices: 60

 

 

 

 

 

 

 

 

 

 

 

Contrast:

FA: 90°
TR/TE: 6614/81 msec
NSA: 1
Fat Suppression: SPIR
Max B-Factor: 1000
Shim: PB Volume (i.e. highest order shims possible within one vol)

 

 

 

 

 

 

Bvalue: 1000

For both the HARDI (b=3000) and DTI (b=1000) acquisitions the gradient tables are used to generate the diffusion gradient directions were constructed using the approach of proposed by Cook et al. for optimal acquisition orders of diffusion-weighted MRI measurements(1)as outlined in their 2007 paper in J. Magn. Res. Imag.  The gradient table of the 68 directions was created with the software ‘camino’, which can be found online at: http://cmic.cs.ucl.ac.uk/camino.  This method for generating the DTI vector is designed to optimize the ordering of gradient directions in diffusion-weighted MRI so that partial scans have the best spherical coverage. Diffusion-weighted MRI often uses a spherical sampling scheme, which acquires images sequentially with diffusion-weighting gradients in unique directions distributed isotropically on the hemisphere. If not all of the measurements can be completed; the quality of diffusion tensors fitted to the partial scan is sensitive to the order of the gradient directions in the scanner protocol. If the directions are in a random order, then a partial scan may cover some parts of the hemisphere densely but other parts sparsely and thus provide poor spherical coverage. The order used here improves the spherical coverage of partial scans and has the advantage of maintaining the optimal coverage of the complete scan.  This approach was thought be appropriate for use in children who might not be able to remain still during an entire 68 direction DTI scan.  Note that 7 b=0 diffusion gradients are interleaved with the directions scans as shown. 

Cook PA, Symms M, Boulby PA, Alexander DC. Optimal acquisition orders of diffusion-weighted MRI measurements. J Magn Reson Imaging. 2007;25(5):1051-8. doi: 10.1002/jmri.20905. PubMed PMID: 17457801. DTI Directions: 68 (61 dirs + 7 B0s)

DTI Gradient Table for 61 Directions plus 7 B=0 images

0 -0 0
0.71109 0.52107 0.47206
0.826 -0.104 0.554
0.23604 -0.80915 0.5381
0.66773 0.71271 0.21491
-0.024995 -0.9338 0.35692
-0.52923 0.78034 0.33314
-0.76884 -0.15697 0.61987
0.81113 0.22203 0.54108
0 -0 0
0.2569 0.58576 0.76869
-0.78381 0.16796 0.59786
0.9225 -0.26014 0.28516
0.15601 0.83604 0.52602
-0.063005 0.69105 0.72005
-0.7029 0.50793 0.49793
0.46988 -0.59085 0.65583
-0.86508 -0.37603 0.33203
-0.37903 -0.89208 -0.24602
0 -0 0
0.34893 -0.065987 0.93482
0.046983 0.96764 0.24791
-0.20002 0.86508 0.46004
-0.647 -0.682 0.341
0.035992 0.055988 0.99778
0.25591 0.96465 -0.062977
-0.19801 -0.49501 0.84602
0.5552 -0.74027 0.37914
0.61569 0.082958 0.78361
0 -0 0
-0.31309 0.93827 0.14704
0.58012 -0.81417 0.024005
0.54729 0.41222 0.72839
0.33594 -0.37994 0.86185
-0.35678 -0.85746 0.37077
-0.55676 0.0079965 0.83064
0.73921 -0.45213 0.49914
0.61213 -0.25605 0.74816
-0.67292 -0.46295 0.57693
0 -0 0
0.025007 0.38811 0.92127
-0.78013 0.5951 -0.19303
0.95972 0.077977 0.26992
0.11799 -0.60695 0.78593
-0.99667 0.074975 0.031989
-0.27991 0.17094 0.94469
-0.92149 0.38821 0.012006
-0.41598 -0.65097 0.63497
0.88506 0.40803 0.22402
0 -0 0
0.81379 0.57785 -0.061984
-0.93718 -0.044008 0.34607
-0.967 -0.25 0.049
-0.42401 0.66202 0.61802
-0.55731 0.34019 0.75742
-0.32711 0.9173 -0.22707
0.55276 0.83064 -0.066971
-0.27406 0.4751 0.83617
0.043006 -0.28304 0.95814
0 -0 0
-0.11395 -0.77167 0.62573
-0.76667 0.61573 0.18192
-0.25608 -0.15405 0.9543
0.45876 0.71562 0.52672
0.066027 -0.99641 0.053022
-0.49095 -0.32596 0.80791
0.32397 0.25898 0.90993
-0.90024 0.29408 0.32108

 

Physiological Imaging Session

Functional Imaging Session

References  

 

Back to the Index Page.