New hybrid technique for accurate and reproducible quantitation of dynamic contrast-enhanced MRI data

A comprehensive evaluation and impact of normalization of generalized tracer kinetic model parameters to characterize blood-brain-barrier permeability in normal-appearing and tumor tissue regions of patients with glioma

RATIONALE AND OBJECTIVES

To comprehensively evaluate robustness and variations of DCE-MRI derived generalized-tracer-kinetic-model (GTKM) parameters in healthy and tumor tissues and impact of normalization in mitigating these variations on application to glioma.

MATERIALS (PATIENTS) AND METHODS

A retrospective study included pre-operative 31 high-grade-glioma(HGG), 22 low-grade-glioma(LGG) and 33 follow-up data from 10 patients a prospective study with 4 HGG subjects. Voxel-wise GTKM was fitted to DCE-MRI data to estimate K^trans, v_e, v_b. Simulations were used to evaluate noise sensitivity. Variation of parameters with-respect-to arterial-input-function (AIF) variation and data length were studied. Normalization of parameters with-respect-to mean values in gray-matter (GM) and white-matter (WM) regions (GM-Type-2, WM-Type-2) and mean curves (GM-Type-1, WM-Type-1) were also evaluated. Co-efficient-of-variation(CoV), relative-percentage-error (RPE), Box-Whisker plots, bar graphs and t-test were used for comparison.

RESULTS

GTKM was fitted well in all tissue regions. K^trans and v_e in contrast-enhancing (CE) has shown improved noise sensitivity in longer data. v_b was reliable in all tissues. Mean AIF and C(t) peaks showed ~38% and ~35% variations. During simulation, normalizations have mitigated variations due to changes in AIF amplitude in K^trans and v_b.. v_e was less sensitive to normalizations. CoV of K^trans and v_b has reduced ~70% after GM-Type-1 normalization and ~80% after GM-Type-2 normalization, respectively. GM-Type-1 (p = 0.003) and GM-Type-2 (p = 0.006) normalizations have significantly improved differentiation of HGG and LGG using K^trans.

CONCLUSION

K^trans and v_b can be reliably estimated in normal-appearing brain tissues and can be used for normalization of corresponding parameters in tumor tissues for mitigating inter-subject variability due to errors in AIF. Normalized K^trans and v_b provided improved differentiation of HGG and LGG.

Intracranial Distribution of Intravenously Administered Gadolinium-based Contrast Agent over a Period of 24 Hours: Evaluation with 3D-real IR Imaging and MR Fingerprinting

PURPOSE

To evaluate the feasibility for the detection of slight contrast effects after intravenous administration of single dose gadolinium-based contrast agent (IV-SD-GBCA), the time course of the GBCA distribution up to 24 h was examined in various fluid spaces and brain parenchyma using 3D-real IR imaging and MR fingerprinting (MRF).

METHODS

Twenty-four patients with a suspicion of endolymphatic hydrops were scanned at pre-administration and at 10 min, 4 and 24 h post-IV-SD-GBCA. 3D-real IR images and MRF at the level of the internal auditory canal were obtained. The signal intensity on the 3D-real IR image of the cerebrospinal fluid (CSF) in the cerebellopontine angle cistern (CPA), Sylvian fissure (Syl), lateral ventricle (LV), and cochlear perilymph (CPL) was measured. The T₁ and T₂ values of cerebellar gray (GM) and white matter (WM) were measured using MRF. Each averaged value at the various time points was compared using an analysis of variance.

RESULTS

The signal intensity on the 3D-real IR image in each CSF region peaked at 4 h, and was decreased significantly by 24 h (P< 0.05). All patients had a maximum signal intensity at 4 h in the CPA, and Syl. The mean CPL signal intensity peaked at 4 h and decreased significantly by 24 h (P < 0.05). All patients but two had a maximum signal intensity at 4 h. Regarding the T₁ value in the cerebellar WM and GM, the T₁ value at 10 min post-IV-GBCA was significantly decreased compared to the pre-contrast scan, but no significant difference was observed at the other time points. There was no significant change in T₂ in the gray or white matter at any of the time points.

CONCLUSION

Time course of GBCA after IV-SD-GBCA could be evaluated by 3D-real IR imaging in CSF spaces and in the brain by MRF.

Blood-brain barrier permeability of normal-appearing white matter in patients with vestibular schwannoma: A new hybrid approach for analysis of T1 -W DCE-MRI

PURPOSE

To develop and assess a "hybrid" method that combines a first-pass analytical approach and the Patlak plot (PP) to improve assessment of low blood-brain barrier permeability from dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) data.

MATERIALS AND METHODS

Seven patients with vestibular schwannoma were enrolled. T1 -W DCE imaging was acquired on a 1.5T scanner. Normal-appearing white matter (NAWM) was divided into four regions of interest (ROIs) based on the magnitude of changes in longitudinal relaxation rate (ΔR1) after gadolinium administration. Kinetic analysis of ROI-averaged contrast agent concentration curves was performed using both the conventional PP and the hybrid method. Computer simulated uptake curves that resemble those from NAWM were analyzed with both methods. Percent deviations (PD) of the "measured" values from the "true" values were calculated to evaluate accuracy and precision of the two methods.

RESULTS

The simulation showed that, at a noise level of 4% (a noise level similar to the in vivo data) and using a signal intensity (SI) averaging scheme, the new hybrid method achieved a PD of 0.9 ± 2.7% for vp , and a PD of -5.4 ± 5.9% for Ktrans . In comparison, the PP method obtained a PD of 3.6 ± 11.3% for vp , and -8.3 ± 12.8% for Ktrans . One-way analyses of variance (ANOVAs) showed significant variations from the four WM regions (P < 10-15 for ΔR1; P < 10-6 for Ktrans ; P < 10-4 for vp ).

CONCLUSION

Both computer simulation and in vivo studies demonstrate improved reliability in vp and Ktrans estimates with the hybrid method.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 1 J. MAGN. RESON. IMAGING 2017;46:79-93.

Analysis of Multicontrast Carotid Plaque MR Imaging

Plaque imaging by MR imaging provides a wealth of information on the characteristics of individual plaque that may reveal vulnerability to rupture, likelihood of progression, or optimal treatment strategy. T1-weighted and T2-weighted images among other options reveal plaque morphology and composition. Dynamic contrast-enhanced-MR imaging reveals plaque activity. To extract this information, image processing tools are needed. Numerous approaches for analyzing such images have been developed, validated against histologic gold standards, and used in clinical studies. These efforts are summarized in this article.

Semi-quantitative parameter analysis of DCE-MRI revisited: monte-carlo simulation, clinical comparisons, and clinical validation of measurement errors in patients with type 2 neurofibromatosis

PURPOSE

To compare semi-quantitative (SQ) and pharmacokinetic (PK) parameters for analysis of dynamic contrast enhanced MR data (DCE-MRI) and investigate error-propagation in SQ parameters.

METHODS

Clinical data was collected from five patients with type 2-neurofibromatosis (NF2) receiving anti-angiogenic therapy for rapidly growing vestibular schwannoma (VS). There were 7 VS and 5 meningiomas. Patients were scanned prior to therapy and at days 3 and 90 of treatment. Data was collected using a dual injection technique to permit direct comparison of SQ and PK parameters. Monte Carlo modeling was performed to assess potential measurement errors in SQ parameters in persistent, washout, and weakly enhancing tissues. The simulation predictions for five semi-quantitative parameters were tested using the clinical DCE-MRI data.

RESULTS

In VS, SQ parameters and Ktrans showed close correlation and demonstrated similar therapy induced reductions. In meningioma, only the denoised Signal Enhancement Ratio (Rse1/se2(DN)) showed a significant therapy induced reduction (p<0.05). Simulation demonstrated: 1) Precision of SQ metrics normalized to the pre-contrast-baseline values (MSErel and ∑MSErel) is improved by use of an averaged value from multiple baseline scans; 2) signal enhancement ratio Rmse1/mse2 shows considerable susceptibility to noise; 3) removal of outlier values to produce a new parameter, Rmse1/mse2(DN), improves precision and sensitivity to therapy induced changes. Direct comparison of in-vivo analysis with Monte Carlo simulation supported the simulation predicted error distributions of semi-quantitative metrics.

CONCLUSION

PK and SQ parameters showed similar sensitivity to anti-angiogenic therapy induced changes in VS. Modeling studies confirmed the benefits of averaging baseline signal from multiple images for normalized SQ metrics and demonstrated poor noise tolerance in the widely used signal enhancement ratio, which is corrected by removal of outlier values.

Repeatability of quantitative MRI measurements in normal breast tissue

PURPOSE

To evaluate the variability and repeatability of repeated magnetic resonance imaging (MRI) measurements in normal breast tissues between and within subjects.

METHODS

Eighteen normal premenopausal subjects underwent two contrast-enhanced MRI scans within 72 hours or during the same menstrual phase in two consecutive months. A subset of nine women also completed diffusion-weighted imaging (DWI). Fibroglandular tissue (FGT) density and FGT enhancement were measured on the contrast-enhanced MRI. Apparent diffusion coefficient (ADC) values were computed from DWI. Between- and within-subject coefficients of variation (bCV and wCV, respectively) were assessed. Repeatability of all measurements was assessed by the coefficient of repeatability (CR) and Bland-Altman plots.

RESULTS

The bCV of FGT density and FGT enhancement at visit 1 and visit 2 ranged from 47% to 63%. The wCV was 13% for FGT density, 22% for FGT enhancement, and 11% for ADC. The CRs of FGT density and FGT enhancement were 0.15 and 0.19, respectively, and for ADC, it was 6.1 x 10(-4) mm(2)/s.

CONCLUSIONS

We present an estimate of the variability and repeatability of MR measurements in normal breasts. These estimates provide the basis for understanding the normal variation of healthy breast tissue in MRI and establishing thresholds for agreement between measurements.

Model selection in measures of vascular parameters using dynamic contrast-enhanced MRI: experimental and clinical applications

A review of the selection of models in dynamic contrast-enhanced MRI (DCE-MRI) is conducted, with emphasis on the balance between the bias and variance required to produce stable and accurate estimates of vascular parameters. The vascular parameters considered as a first-order model are the forward volume transfer constant K(trans) , the plasma volume fraction vp and the interstitial volume fraction ve . To illustrate the critical issues in model selection, a data-driven selection of models in an animal model of cerebral glioma is followed. Systematic errors and extended models are considered. Studies with nested and non-nested pharmacokinetic models are reviewed; models considering water exchange are considered.

The effect of blood inflow and B(1)-field inhomogeneity on measurement of the arterial input function in axial 3D spoiled gradient echo dynamic contrast-enhanced MRI

A major potential confound in axial 3D dynamic contrast-enhanced magnetic resonance imaging studies is the blood inflow effect; therefore, the choice of slice location for arterial input function measurement within the imaging volume must be considered carefully. The objective of this study was to use computer simulations, flow phantom, and in vivo studies to describe and understand the effect of blood inflow on the measurement of the arterial input function. All experiments were done at 1.5 T using a typical 3D dynamic contrast-enhanced magnetic resonance imaging sequence, and arterial input functions were extracted for each slice in the imaging volume. We simulated a set of arterial input functions based on the same imaging parameters and accounted for blood inflow and radiofrequency field inhomogeneities. Measured arterial input functions along the vessel length from both in vivo and the flow phantom agreed with simulated arterial input functions and show large overestimations in the arterial input function in the first 30 mm of the vessel, whereas arterial input functions measured more centrally achieve accurate contrast agent concentrations. Use of inflow-affected arterial input functions in tracer kinetic modeling shows potential errors of up to 80% in tissue microvascular parameters. These errors emphasize the importance of careful placement of the arterial input function definition location to avoid the effects of blood inflow.

Kinetic assessment of breast tumors using high spatial resolution signal enhancement ratio (SER) imaging

The goal of this study was to investigate the relationship between an empirical contrast kinetic parameter, the signal enhancement ratio (SER), for three-timepoint, high spatial resolution contrast-enhanced (CE) MRI, and a commonly analyzed pharmacokinetic parameter, kep, using dynamic high temporal resolution CE-MRI. Computer simulation was performed to investigate: 1) the relationship between the SER and the contrast agent concentration ratio (CACR) of two postcontrast timepoints (tp1 and tp2); 2) the relationship between the CACR and the redistribution rate constant (kep) based on a two-compartment pharmacokinetic model; and 3) the sensitivity of the relationship between the SER and kep to native tissue T1 relaxation time, T10, and to errors in an assumed vascular input function. The relationship between SER and kep was verified experimentally using a mouse model of breast cancer. The results showed that a monotonic mathematical relationship between SER and kep could be established if the acquisition parameters and the two postinjection timepoints of SER, tp1, tp2, were appropriately chosen. The in vivo study demonstrated a close correlation between SER and kep on a pixel-by-pixel basis (Spearman rank correlation coefficient=0.87+/-0.03). The SER is easy to calculate and may have a unique role in breast tissue characterization.

Imaging tumor vascular heterogeneity and angiogenesis using dynamic contrast-enhanced magnetic resonance imaging

This article reviews the application of dynamic contrast-enhanced magnetic resonance imaging in both clinical studies and early-phase trials of angiogenesis inhibitors. Emphasis is placed on how variation in image acquisition and analysis affects the meaning and use of derived variables. We then review the potential for future developments, with particular reference to the application of dynamic contrast-enhanced magnetic resonance imaging to evaluate the heterogeneity of tumor tissues.

Comparison of cerebral blood volume maps generated fromT2* andT1weighted MRI data in intra-axial cerebral tumours

We compared parametric maps, measured values and value distributions of cerebral blood volume (CBV) derived from (1) first pass T1 weighted dynamic contrast-enhanced (DCE) data (T1-CBV) using the recently described leakage profile model and (2) conventional T2* weighted DCE data (T2*-CBV) using a conventional curve fitting technique, in nine patients with intraaxial tumours. Regions of interest were defined around enhancing tumour tissue on matched slices. Median tumour values and conspicuity indexes of CBV from the two techniques were compared, demonstrating good correlation (r = 0.667,p<0.05) in enhancing tumour and no significant difference in conspicuity. Pixel-by-pixel scattergrams of values in normal brain in a representative matched slice were produced for each case, which showed excellent correlation (r = 0.96,p<0.001). Distortion of blood vessels around susceptibility interfaces was evident on T2* CBV but not on T1 CBV maps. Leakage-free T1 CBV maps do not suffer from the susceptibility artifacts seen in T2* CBV maps, although they present comparable biological information.

Experimentally-derived functional form for a population-averaged high-temporal-resolution arterial input function for dynamic contrast-enhanced MRI

Rapid T(1)-weighted 3D spoiled gradient-echo (GRE) data sets were acquired in the abdomen of 23 cancer patients during a total of 113 separate visits to allow dynamic contrast-enhanced MRI (DCE-MRI) analysis of tumor microvasculature. The arterial input function (AIF) was measured in each patient at each visit using an automated AIF extraction method following a standardized bolus administration of gadodiamide. The AIFs for each patient were combined to obtain a mean AIF that is representative for any individual. The functional form of this general AIF may be useful for studies in which AIF measurements are not possible. Improvements in the reproducibility of DCE-MRI model parameters (K(trans), v(e), and v(p)) were observed when this new, high-temporal-resolution population AIF was used, indicating the potential for increased sensitivity to therapy-induced change.

Capillary physiology of human medulloblastoma: impact on chemotherapy

BACKGROUND

Advances in the treatment of medulloblastoma have largely been attributed to the introduction of chemotherapy, although Phase III trials have shown advantages for chemotherapy only in subgroups. Because the efficacy of chemotherapy depends on tumor vascularization, the vascular physiology of human medulloblastomas was evaluated.

METHODS

Seven patients with histologically proven medulloblastomas underwent measurements of capillary permeability and vascular plasma volume using contrast-enhanced dynamic computer tomography. Regional blood flow was measured in 5 patients using xenon computed tomography (CT).

RESULTS

The capillary permeability-surface product for water-soluble compounds ranged from 1.7 +/- 5.5 to 17.6 +/- 12.3 muL/g/min with a mean of 10.5 +/- 6.3 microL/g/min. The vascular plasma volume ranged from 0.02 +/- 0.021 to 0.045 +/- 0.049 mL/g with a mean of 0.03 +/- 0.01 mL/g. The efflux rate ranged from 0.012 +/- 0.007 to 0.065 +/- 0.064 1/min with a mean of 0.039 +/- 0.020 1/min. Regional tumoral blood flow showed a mean of 19.86 +/- 6.8 mL/100g/min as compared with normal cerebellum with 45.4 +/- 12.03 mL/100g/min (P < .005).

CONCLUSIONS

The current study demonstrated a low capillary permeability and blood flow in medulloblastomas that could explain the limited response rates of partially resected tumors even after aggressive high-dose chemotherapy, as recently reported.

Estimation of contrast agent concentration in intra- and extra-vascular spaces of brain tissue

This article presents a new method for estimating the leakage of a contrast agent out of a vessel. The proposed method is developed based on tissue homogeneity (TH) model, modified Patlak model, and Monte Carlo simulation. The analytical methods published in the literature estimate the contrast agent leakage by solving the coupled differential equations associated with the TH model under adiabatic conditions. These methods employ unrealistic simplifying assumptions and become intractable in their applications to the vessels that have a non-uniform permeability. Without making any unrealistic assumptions, our approach simply tracks the passage of the contrast agent through the capillary and its crossing of the vessel walls based on the blood flow in the vessel, the vessel's permeability, and the condition of the blood-brain barrier (BBB). These are treated as statistical processes that can be modeled reasonably well using the Monte Carlo method. In the proposed approach, the intra- and extra-vascular spaces are divided into multiple compartments, similar to the Patlak model. A real, measured arterial input function (AIF) is used as the capillary input and the concentration of the contrast agent is found as a function of time and distance, inside and outside of the capillary. This is done for normal and abnormal capillaries with uniform and non-uniform permeability. The proposed method generates concentration curves similar to those of the analytical method for simple AIF models. It also generates reasonable concentration curves for a real AIF. The proposed method does not fit a mathematical function to the measured AIF and does not make unrealistic simplifying assumptions. It is not therefore prone to the fitting errors and generates more realistic and more accurate results than the analytical methods.

Comparative study of methods for determining vascular permeability and blood volume in human gliomas

PURPOSE

To characterize human gliomas using T1-weighted dynamic contrast-enhanced MRI (DCE-MRI), and directly compare three pharmacokinetic analysis techniques: a conventional established technique and two novel techniques that aim to reduce erroneous overestimation of the volume transfer constant between plasma and the extravascular extracellular space (EES) (Ktrans) in areas of high blood volume.

MATERIALS AND METHODS

Eighteen patients with high-grade gliomas underwent DCE-MRI. Three kinetic models were applied to estimate Ktrans and fractional blood plasma volume (vp). We applied the Tofts and Kermode (TK) model without arterial input function (AIF) estimation, the TK model modified to include vp and AIF estimation (mTK), and a "first pass" variant of the TK model (FP).

RESULTS

KTK values were considerably higher than KmTK and KFP values (P <0.001). KmTK and KFP were more comparable and closely correlated (rho=0.744), with KmTK generally higher than KFP (P <0.001). Estimates of vp(mTK) and vp(FP) also showed a significant difference (P <0.001); however, these values were very closely correlated (rho=0.901). KTK parameter maps showed "pseudopermeability" effects displaying numerous vessels. These were not visualized on KmTK and KFP maps but appeared on the corresponding vp maps, indicating a failure of the TK model in commonly occurring vascular regions.

CONCLUSION

Both of the methods that incorporate a measured AIF and an estimate of vp provide similar pathophysiological information and avoid erroneous overestimation of Ktrans in areas of significant vessel density, and thus allow a more accurate estimation of endothelial permeability.

A comparison of Ktransmeasurements obtained with conventional and first pass pharmacokinetic models in human gliomas

PURPOSE

To compare in a group of patients with cerebral gliomas the estimates of Ktrans between a conventionally established pharmacokinetic model and a recently developed first pass method.

MATERIALS AND METHODS

Glioma patients (23) were studied using T1-weighted dynamic contrast-enhanced magnetic resonance imaging (MRI), and two alternative pharmacokinetic models were used for analysis to derive the volume transfer constant Ktrans. These were a modified version of the established model (yielding KTK) and a recently published method based on first pass leakage profile (FP) of contrast bolus (yielding Kfp).

RESULTS

We found a strong correlation between intra-tumoral median KTK and Kfp (rho = 0.650, P < 0.01), but the values from the conventional model were consistently and significantly higher (mean of inter-tumoral Kfp and KTK medians were 0.018 minute(-1) and 0.284 minute(-1), respectively, P < 0.001). The spatial distribution of KTK and Kfp showed poor correlation in the presence of large vascular structures and good correlation elsewhere.

CONCLUSION

KTK and Kfp produce similar biologic information within voxels not dominated by vascular tissue. The FP method avoids erroneous overestimation of Ktrans in areas of significant intravascular contrast. Findings are in keeping with the predictions of previous mathematical simulations.