Absolute quantification of perfusion using dynamic susceptibility contrast MRI: pitfalls and possibilities

Absolute quantification of cerebral blood flow, cerebral blood volume and mean transit time is desirable in the determination of tissue viability thresholds and tissue at risk in acute ischaemic stroke, as well as in cases where a global reduction in cerebral blood flow is expected, for example, in patients with dementia or depressive disorders. Absolute values are also useful when comparing sequential examinations of tissue perfusion parameters, for example, in the monitoring and follow-up of various kinds of therapy. Regardless of the method employed, a number of assumptions and approximations must be made to obtain absolute measures of perfusion. Furthermore, the different stages of data acquisition and processing are associated with various degrees of uncertainty. In this review, the problems of particular relevance to absolute quantification of cerebral perfusion parameters using dynamic susceptibility contrast magnetic resonance imaging are discussed, and possible solutions are outlined.     

Denoising of arterial spin labeling data: wavelet-domain filtering compared with Gaussian smoothing

Purpose  

To investigate a wavelet-based filtering scheme for denoising of arterial spin labeling (ASL) data, potentially enabling reduction of the required number of averages and the acquisition time.     

Non-parametric deconvolution using Bézier curves for quantification of cerebral perfusion in dynamic susceptibility contrast MRI

Objective

Deconvolution is an ill-posed inverse problem that tends to yield non-physiological residue functions R(t) in dynamic susceptibility contrast magnetic resonance imaging (DSC-MRI). In this study, the use of Bézier curves is proposed for obtaining physiologically reasonable residue functions in perfusion MRI.

Materials and methods

Cubic Bézier curves were employed, ensuring R(0) = 1, bounded-input, bounded-output stability and a non-negative monotonically decreasing solution, resulting in 5 parameters to be optimized. Bézier deconvolution (BzD), implemented in a Bayesian framework, was tested by simulation under realistic conditions, including effects of arterial delay and dispersion. BzD was also applied to DSC-MRI data from a healthy volunteer.

Results

Bézier deconvolution showed robustness to different underlying residue function shapes. Accurate perfusion estimates were observed, except for boxcar residue functions at low signal-to-noise ratio. BzD involving corrections for delay, dispersion, and delay with dispersion generally returned accurate results, except for some degree of cerebral blood flow (CBF) overestimation at low levels of each effect. Maps of mean transit time and delay were markedly different between BzD and block-circulant singular value decomposition (oSVD) deconvolution.

Discussion

A novel DSC-MRI deconvolution method based on Bézier curves was implemented and evaluated. BzD produced physiologically plausible impulse response, without spurious oscillations, with generally less CBF underestimation than oSVD.

     

Correlation between arterial blood volume obtained by arterial spin labelling and cerebral blood volume in intracranial tumours

Objective  

To compare measurements of the arterial blood volume (aBV), a perfusion parameter calculated from arterial spin labelling (ASL), and cerebral blood volume (CBV), calculated from dynamic susceptibility contrast (DSC) MRI. In the clinic, CBV is used for grading of intracranial tumours.     

Quantitative diffusion coefficient maps using fast spin-echo MRI

In this work, we have evaluated the performance of a diffusion-sensitive fast spin-echo (FSE) pulse sequence. The proposed pulse sequence utilises velocity-compensating diffusion-encoding gradients and includes the collection of navigator echoes. Spoiler gradients were inserted in the slice-selecting direction to minimise effects from stimulated echoes. Calculations of the b values showed that cross-terms between imaging gradients and diffusion gradients only led to a marginal increase of b values. Pixel-wise calculation of apparent diffusion coefficient (ADC) maps was performed numerically, considering cross-terms between diffusion-encoding and imaging gradients. The sequences investigated used echo train lengths of 16, 8 and 4 echoes and were encoded in either the slice-, frequency- or phase-encoding direction. In order to allow for higher b values a pulse-sequence version using non-motion compensating diffusion-encoding gradients was written. Phantom measurements were performed and the diffusion coefficients of water and acetone were reasonable. Seven healthy volunteers (age 28-50 years) were examined and apparent diffusion coefficient values agreed well with expected values. Diffusion-weighted images, apparent diffusion coefficient maps and images corresponding to the trace of the diffusion tensor of good quality were retrieved in vivo.     

Accuracy of q-space related parameters in MRI: simulations and phantom measurements

The accuracy of q-space measurements was evaluated at a 3.0-T clinical magnetic resonance imaging (MRI) scanner, as compared with a 4.7-T nuclear magnetic resonance (NMR) spectrometer. Measurements were performed using a stimulated-echo pulse-sequence on n-decane as well as on polyethylene glycol (PEG) mixed with different concentrations of water, in order to obtain bi-exponential signal decay curves. The diffusion coefficients as well as the modelled diffusional kurtosis K(fit) were obtained from the signal decay curve, while the full-width at half-maximum (FWHM) and the diffusional kurtosis K were obtained from the displacement distribution. Simulations of restricted diffusion, under conditions similar to those obtainable with a clinical MRI scanner, were carried out assuming various degrees of violation of the short gradient pulse (SGP) condition and of the long diffusion time limit. The results indicated that an MRI system can not be used for quantification of structural sizes less than about 10 microm by means of FWHM since the parameter underestimates the confinements due to violation of the SGP condition. However, FWHM can still be used as an important contrast parameter. The obtained kurtosis values were lower than expected from theory and the results showed that care must be taken when interpreting a kurtosis estimate deviating from zero.     

Absolute cerebral blood flow measured by dynamic susceptibility contrast MRI: a direct comparison with Xe-133 SPECT

Absolute regional cerebral blood flow (CBF) was measured in ten healthy volunteers, using both dynamic susceptibility-contrast (DSC) magnetic resonance imaging (MRI) and Xe-133 SPECT within 4 h. After i.v. injection of Gd-DTPA-BMA (0.3 mmol/kg b.w.), the bolus was monitored with a Simultaneous Dual FLASH pulse sequence (1.5 s/image), providing one slice through brain tissue and a second slice through the carotid artery. Concentration C(t) is proportional to -(1/TE) ln[S(t)/S(0)] was related to CBF as C(t) = CBF [AIF(t) x R(t)], where AIF is the arterial input function and R(t) is the residue function. A singular-value-decomposition-based deconvolution technique was used for retrieval of R(t). Absolute CBF was given by Zierler's area-to-height relation and the central volume principle. For elimination of large vessels (ELV), all MRI-based CBF values exceeding 2.5 times the mean CBF value of the slice were excluded. A correction for partial-volume effects (CPVE) in the artery used for AIF monitoring was based on registration of signal in a phantom with tubes of various diameters (1.5-6.5 mm), providing an individual concentration correction factor applied to AIF data registered in vivo. In the Xe-133 SPECT investigation, 3,000-4,000 MBq of Xe-133 was administered intravenously, and CBF was calculated using the Kanno Lassen algorithm. When ELV and CPVE were applied, DSC-MRI showed average CBF values from the entire slice of 43 +/- 10 ml/(min 100 g) (small-artery AIF) and 48 +/- 17 ml/(min 100 g) (carotid-artery AIF) (mean +/- S.D., n = 10). The corresponding Xe-133-SPECT-based CBF was 33 +/- 6 ml/(min 100 g) (n = 10). The relationships of CBF(MRI) versus CBF(SPECT) showed good linear correlation (r = 0.74-0.83).     

Wavelet-based noise reduction for improved deconvolution of time-series data in dynamic susceptibility-contrast MRI

Dynamic susceptibility-contrast (DSC) MRI requires deconvolution to retrieve the tissue residue function R(t) and the cerebral blood flow (CBF). In this study, deconvolution of time-series data was performed by wavelet-transform-based denoising combined with the Fourier transform (FT). Traditional FT-based deconvolution of noisy data requires frequency-domain filtering, often leading to excessive smoothing of the recovered signal. In the present approach, only a low degree of regularisation was employed while the major noise reduction was accomplished by wavelet transformation of data and Wiener-like filtering in the wavelet space. After inverse wavelet transform, the estimate of CBF·R(t) was obtained. DSC-MRI signal-versus-time curves (signal-to-noise ratios 40 and 100) were simulated, corresponding to CBF values in the range 10–60 ml/(min 100 g). Three shapes of the tissue residue function were investigated. The technique was also applied to six volunteers. Simulations showed CBF estimates with acceptable accuracy and precision, as well as independence of any time shift between the arterial input function and the tissue concentration curve. The grey-matter to white-matter CBF ratio in volunteers was 2.4±0.2. The proposed wavelet/FT deconvolution is robust and can be implemented into existing perfusion software. CBF maps from healthy volunteers showed high quality.     

Dimensionality reduction of fMRI time series data using locally linear embedding

Objective

Data-driven methods for fMRI analysis are useful, for example, when an a priori model of signal variations is unavailable. However, activation sources are typically assumed to be linearly mixed, although non-linear properties of fMRI data, including resting-state data, have been observed. In this work, the non-linear locally linear embedding (LLE) algorithm is introduced for dimensionality reduction of fMRI time series data.

Materials and methods

LLE performance was optimised and tested using simulated and volunteer data for task-evoked responses. LLE was compared with principal component analysis (PCA) as a preprocessing step to independent component analysis (ICA). Using an example data set with known non-linear properties, LLE-ICA was compared with PCA-ICA and non-linear PCA-ICA. A resting-state data set was analysed to compare LLE-ICA and PCA-ICA with respect to identifying resting-state networks.

Results

LLE consistently found task-related components as well as known resting-state networks, and the algorithm compared well to PCA. The non-linear example data set demonstrated that LLE, unlike PCA, can separate non-linearly modulated sources in a low-dimensional subspace. Given the same target dimensionality, LLE also performed better than non-linear PCA.

Conclusion

LLE is promising for fMRI data analysis and has potential advantages compared with PCA in terms of its ability to find non-linear relationships.     

Dynamic susceptibility contrast perfusion MRI using phase-based venous output functions: comparison with pseudo-continuous arterial spin labelling and assessment of contrast agent concentration in large veins

Objectives

Contrast agent (CA) relaxivities are generally not well established in vivo, and the relationship between frequency/phase shift and magnetic susceptibility might be a useful alternative for CA quantification.

Materials and methods

Twenty volunteers (25–84 years old) were investigated using test–retest pre-bolus dynamic susceptibility-contrast (DSC) magnetic resonance imaging (MRI). The pre-bolus phase-based venous output function (VOF) time integral was used for arterial input function (AIF) rescaling. Resulting cerebral blood flow (CBF) data for grey matter (GM) were compared with pseudo-continuous arterial spin labelling (ASL). During the main bolus CA passage, the apparent spatial shift (pixel shift) of the superior sagittal sinus (seen in single-shot echo-planar imaging (EPI)) was converted to CA concentration and compared with conventional ΔR2*-based data and with a predicted phase-based VOF from the pre-bolus experiment.

Results

The phase-based pre-bolus VOF resulted in a reasonable inter-individual GM CBF variability (coefficient of variation 28 %). Comparison with ASL CBF values implied a tissue R2*-relaxivity of 32 mM^?1 s^?1. Pixel-shift data at low concentrations (data not available at peak concentrations) were in reasonable agreement with the predicted phase-based VOF.

Conclusion

Susceptibility-induced phase shifts and pixel shifts are potentially useful for large-vein CA quantification. Previous predictions of a higher R2*-relaxivity in tissue than in blood were supported.