Improving MR quantification of regional blood volume with intravascular T₁ contrast agents: Accuracy, precision, and water exchange

Abstract

The goal of this work was to develop a comprehensive understanding of the relationship between vascular proton exchange rates and the accuracy and precision of tissue blood volume estimates using intravascular T1 contrast agents. Using computer simulations, the effects of vascular proton exchange and experimental pulse sequence parameters on measurement accuracy were quantified. T1 and signal measurements made in a rat model implanted with R3230 mammary adenocarcinoma tumors demonstrated that the theoretical findings are biologically relevant; data demonstrated that over-simplified exchange models may result in measures of tumor, muscle, and liver blood volume fractions that depend on experimental parameters such as the vascular contrast concentration. As a solution to the measurement of blood volume in tissues with exchange that is unknown, methods that minimize exchange rate dependence were examined. Simulations that estimated both the accuracy and precision of such methods indicated that both the inversion recovery and the transverse-spoiled gradient echo methods using a "no-exchange" model provide the best trade-off between accuracy and precision.

T1 measurement of flowing blood and arterial input function determination for quantitative 3D T1‐weighted DCE‐MRI

“…Corrected v b (2.47 Ϯ 0.65%) agreed with other reports, especially those in rabbit tibialis anterior (2.1 Ϯ 0.7%) (14), the most similar to the muscle analyzed in our study. It is noteworthy that our v b agreed with those obtained using entirely different methods, such as a steady-state MRI (1.5 Ϯ 1.0%) (16) or autoradiography technique (2.29 Ϯ 0.28%) (17). Despite the use of corrected AIFs, our v b results may be slightly lower than the truth, due to potential effects of water exchange and since the true v b is known to be an upper bound on Patlak's estimate (18).…”

“…If exchange were slow to intermediate, T 1 relaxation becomes multiexponential, reflecting distinct relaxations in different compartments. The fast exchange assumption is supported by other findings for interstitial-intracellular water movement (16). Departure from this assumption, if present, would have occurred across the vascular-interstitial barrier.…”

“…Agreement was found, particularly with AIFs derived from corrected preinjection blood T 1 . Figure 5 compares this study to literature values for muscle (14,16,17). Better agreement was obtained for parameters derived from corrected AIFs.…”

“…However, we minimized the effects of exchange by maintaining low blood contrast concentrations using: 1) a low injection dose and 2) relatively slow AIF sampling to avoid high transient concentrations during first-pass. In the worst case, the effect of limited exchange on DCE-MRI analysis is underestimation of v b and overestimation of K trans (16,18). However, at least for K trans , the effects of exchange were avoided by using Patlak's model (18).…”

Absolute quantification of perfusion by dynamic susceptibility contrast MRI using Bookend and VASO steady-state CBV calibration: a comparison with pseudo-continuous ASL

“…Eq. 3 assumes fast water exchange between the intra-and the extra-vascular space [14,15], and WCF(ΔR 1 ) corrects for expected deviations from the fast-water-exchange limit [7]. The WCF is dependent on the concentration of CA in blood as well as on sequence parameters in the T1 measurements and can be obtained by fitting a curve, of the form a·ΔR 1 2 +b·ΔR 1 +c to data plotted as a function of the change in longitudinal relaxation rate, ΔR1, in blood [7,16].…”