Limits...
Cerebral blood flow quantification using vessel-encoded arterial spin labeling.

Okell TW, Chappell MA, Kelly ME, Jezzard P - J. Cereb. Blood Flow Metab. (2013)

Bottom Line: Experimental results in healthy volunteers showed that there is no systematic bias in the CBF estimates produced by VEPCASL and that the signal-to-noise ratio of the two techniques is comparable.Although more complex acquisition and image processing is required and the potential for motion sensitivity is increased, VEPCASL provides comparable data to PCASL but with the added benefit of vessel-selective information.This could lead to more accurate CBF estimates in patients with a significant collateral flow.

View Article: PubMed Central - PubMed

Affiliation: Nuffield Department of Clinical Neurosciences, Centre for Functional Magnetic Resonance Imaging of the Brain, University of Oxford, Oxford, UK.

ABSTRACT
Arterial spin labeling (ASL) techniques are gaining popularity for visualizing and quantifying cerebral blood flow (CBF) in a range of patient groups. However, most ASL methods lack vessel-selective information, which is important for the assessment of collateral flow and the arterial supply to lesions. In this study, we explored the use of vessel-encoded pseudocontinuous ASL (VEPCASL) with multiple postlabeling delays to obtain individual quantitative CBF and bolus arrival time maps for each of the four main brain-feeding arteries and compared the results against those obtained with conventional pseudocontinuous ASL (PCASL) using matched scan time. Simulations showed that PCASL systematically underestimated CBF by up to 37% in voxels supplied by two arteries, whereas VEPCASL maintained CBF accuracy since each vascular component is treated separately. Experimental results in healthy volunteers showed that there is no systematic bias in the CBF estimates produced by VEPCASL and that the signal-to-noise ratio of the two techniques is comparable. Although more complex acquisition and image processing is required and the potential for motion sensitivity is increased, VEPCASL provides comparable data to PCASL but with the added benefit of vessel-selective information. This could lead to more accurate CBF estimates in patients with a significant collateral flow.

Show MeSH

Related in: MedlinePlus

Overview of the processing performed to obtain quantitative cerebral blood flow (CBF), CBF uncertainty and bolus arrival time (BAT) maps from the vessel-encoded (VE) and standard pseudocontinuous arterial spin labeling (PCASL) data. The VEPCASL data give rise to one map per feeding artery, but these are shown together here for brevity. Partial volume estimates (PVEs) derived from the structural image were used to generate a gray matter (GM) mask used in subsequent analyses (see Results). Time-series data are represented with three dots. Various FMRIB Software Library (FSL25) tools were used for brain extraction (BET26), linear (FLIRT22) and non-linear (FNIRT27) registration, motion correction (MCFLIRT22), segmentation (FAST28) and non-linear model fitting of ASL data (BASIL29).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3824178&req=5

fig2: Overview of the processing performed to obtain quantitative cerebral blood flow (CBF), CBF uncertainty and bolus arrival time (BAT) maps from the vessel-encoded (VE) and standard pseudocontinuous arterial spin labeling (PCASL) data. The VEPCASL data give rise to one map per feeding artery, but these are shown together here for brevity. Partial volume estimates (PVEs) derived from the structural image were used to generate a gray matter (GM) mask used in subsequent analyses (see Results). Time-series data are represented with three dots. Various FMRIB Software Library (FSL25) tools were used for brain extraction (BET26), linear (FLIRT22) and non-linear (FNIRT27) registration, motion correction (MCFLIRT22), segmentation (FAST28) and non-linear model fitting of ASL data (BASIL29).

Mentions: A schematic of the processing stages is shown in Figure 2. The PCASL and VEPCASL raw images were motion corrected22 using the head coil calibration image as a reference, before averaging across repeats. A pairwise control minus tag subtraction was then performed on the PCASL data to generate a perfusion image at each PLD. The separation of signals arising from each feeding artery in the VEPCASL data was performed using a maximum a posteriori (MAP) solution (method BT3)23 to the general Bayesian framework24 with two vessels per class. This approach can account for rigid subject motion between the planning TOF and VEPCASL acquisitions. For the purposes of the SNR comparison, this processing was repeated with a matrix inversion (MI) approach (see Chappell et al23), assuming no motion between the TOF and VEPCASL acquisitions.


Cerebral blood flow quantification using vessel-encoded arterial spin labeling.

Okell TW, Chappell MA, Kelly ME, Jezzard P - J. Cereb. Blood Flow Metab. (2013)

Overview of the processing performed to obtain quantitative cerebral blood flow (CBF), CBF uncertainty and bolus arrival time (BAT) maps from the vessel-encoded (VE) and standard pseudocontinuous arterial spin labeling (PCASL) data. The VEPCASL data give rise to one map per feeding artery, but these are shown together here for brevity. Partial volume estimates (PVEs) derived from the structural image were used to generate a gray matter (GM) mask used in subsequent analyses (see Results). Time-series data are represented with three dots. Various FMRIB Software Library (FSL25) tools were used for brain extraction (BET26), linear (FLIRT22) and non-linear (FNIRT27) registration, motion correction (MCFLIRT22), segmentation (FAST28) and non-linear model fitting of ASL data (BASIL29).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3824178&req=5

fig2: Overview of the processing performed to obtain quantitative cerebral blood flow (CBF), CBF uncertainty and bolus arrival time (BAT) maps from the vessel-encoded (VE) and standard pseudocontinuous arterial spin labeling (PCASL) data. The VEPCASL data give rise to one map per feeding artery, but these are shown together here for brevity. Partial volume estimates (PVEs) derived from the structural image were used to generate a gray matter (GM) mask used in subsequent analyses (see Results). Time-series data are represented with three dots. Various FMRIB Software Library (FSL25) tools were used for brain extraction (BET26), linear (FLIRT22) and non-linear (FNIRT27) registration, motion correction (MCFLIRT22), segmentation (FAST28) and non-linear model fitting of ASL data (BASIL29).
Mentions: A schematic of the processing stages is shown in Figure 2. The PCASL and VEPCASL raw images were motion corrected22 using the head coil calibration image as a reference, before averaging across repeats. A pairwise control minus tag subtraction was then performed on the PCASL data to generate a perfusion image at each PLD. The separation of signals arising from each feeding artery in the VEPCASL data was performed using a maximum a posteriori (MAP) solution (method BT3)23 to the general Bayesian framework24 with two vessels per class. This approach can account for rigid subject motion between the planning TOF and VEPCASL acquisitions. For the purposes of the SNR comparison, this processing was repeated with a matrix inversion (MI) approach (see Chappell et al23), assuming no motion between the TOF and VEPCASL acquisitions.

Bottom Line: Experimental results in healthy volunteers showed that there is no systematic bias in the CBF estimates produced by VEPCASL and that the signal-to-noise ratio of the two techniques is comparable.Although more complex acquisition and image processing is required and the potential for motion sensitivity is increased, VEPCASL provides comparable data to PCASL but with the added benefit of vessel-selective information.This could lead to more accurate CBF estimates in patients with a significant collateral flow.

View Article: PubMed Central - PubMed

Affiliation: Nuffield Department of Clinical Neurosciences, Centre for Functional Magnetic Resonance Imaging of the Brain, University of Oxford, Oxford, UK.

ABSTRACT
Arterial spin labeling (ASL) techniques are gaining popularity for visualizing and quantifying cerebral blood flow (CBF) in a range of patient groups. However, most ASL methods lack vessel-selective information, which is important for the assessment of collateral flow and the arterial supply to lesions. In this study, we explored the use of vessel-encoded pseudocontinuous ASL (VEPCASL) with multiple postlabeling delays to obtain individual quantitative CBF and bolus arrival time maps for each of the four main brain-feeding arteries and compared the results against those obtained with conventional pseudocontinuous ASL (PCASL) using matched scan time. Simulations showed that PCASL systematically underestimated CBF by up to 37% in voxels supplied by two arteries, whereas VEPCASL maintained CBF accuracy since each vascular component is treated separately. Experimental results in healthy volunteers showed that there is no systematic bias in the CBF estimates produced by VEPCASL and that the signal-to-noise ratio of the two techniques is comparable. Although more complex acquisition and image processing is required and the potential for motion sensitivity is increased, VEPCASL provides comparable data to PCASL but with the added benefit of vessel-selective information. This could lead to more accurate CBF estimates in patients with a significant collateral flow.

Show MeSH
Related in: MedlinePlus