Background: Pharmacokinetic (PK) models can describe microvascular density and integrity. An essential component of PK models is the arterial input function (AIF) representing the time-dependent concentration of contrast agent (CA) in the blood plasma supplied to a tissue. Purpose/Hypothesis: To evaluate a novel method for subject-specific AIF estimation that takes inflow effects into account. Study Type: Retrospective study. Subjects: Thirteen clinical patients referred for spine-related complaints; 21 patients from a study into luminal Crohn's disease with known Crohn's Disease Endoscopic Index of Severity (CDEIS). Field Strength/Sequence: Dynamic fast spoiled gradient echo (FSPGR) at 3T. Assessment: A population-averaged AIF, AIFs derived from distally placed regions of interest (ROIs), and the new AIF method were applied. Tofts' PK model parameters (including v_p and K^trans) obtained with the three AIFs were compared. In the Crohn's patients K^trans was correlated to CDEIS. Statistical Tests: The median values of the PK model parameters from the three methods were compared using a Mann–Whitney U-test. The associated variances were statistically assessed by the Brown-Forsythe test. Spearman's rank correlation coefficient was computed to test the correlation of K^trans to CDEIS. Results: The median v_p was significantly larger when using the distal ROI approach, compared to the two other methods (P < 0.05 for both comparisons, in both applications). Also, the variances in v_p were significantly larger with the ROI approach (P < 0.05 for all comparisons). In the Crohn's disease study, the estimated K^trans parameter correlated better with the CDEIS (r = 0.733, P < 0.001) when the proposed AIF was used, compared to AIFs from the distal ROI method (r = 0.429, P = 0.067) or the population-averaged AIF (r = 0.567, P = 0.011). Data Conclusion: The proposed method yielded realistic PK model parameters and improved the correlation of the K^trans parameter with CDEIS, compared to existing approaches. Level of Evidence: 3. Technical Efficacy Stage 1. J. Magn. Reson. Imaging 2018;47:1197–1204.

@en

Background: The arterial input function (AIF) represents the time-dependent arterial contrast agent (CA) concentration that is used in pharmacokinetic modeling. Purpose: To develop a novel method for estimating the AIF from dynamic contrast-enhanced (DCE-) MRI data, while compensating for flow enhancement. Study Type: Signal simulation and phantom measurements. Phantom Model: Time–intensity curves (TICs) were simulated for different numbers of excitation pulses modeling flow effects. A phantom experiment was performed in which a solution (without CA) was passed through a straight tube, at constant flow velocity. Field Strength/Sequence: Dynamic fast spoiled gradient echo (FSPGRs) at 3T MRI, both in the simulations and in the phantom experiment. TICs were generated for a duration of 373 seconds and sampled at intervals of 1.247 seconds (300 timepoints). Assessment: The proposed method first estimates the number of pulses that spins have received, and then uses this knowledge to accurately estimate the CA concentration. Statistical Tests: The difference between the median of the estimated number of pulses and the true value was deter- mined, as well as the interquartile range (IQR) of the estimations. The estimated CA concentrations were evaluated in the same way. The estimated number of pulses was also used to calculate flow velocity. Results: The difference between the median estimated and reference number of pulses varied from –0.005 to –1.371 (corresponding IQRs: 0.853 and 48.377) at true values of 10 and 180 pulses, respectively. The difference between the median estimated CA concentration and the reference value varied from –0.00015 to 0.00306 mmol/L (corresponding IQRs: 0.01989 and 1.51013 mmol/L) at true values of 0.5 and 8.0 mmol/l, respectively, at an intermediate value of 100 pulses. The estimated flow velocities in the phantom were within 10% of the reference value. Data Conclusion: The proposed method accurately corrects the MRI signal affected by the inflow effect. Level of Evidence: 1 Technical Efficacy: Stage 1 @en

Dynamic Contrast Enhanced MRI is an important technique to assess the pharmacokinetic properties of tissues. This thesis addresses two major steps necessary for quantitative DCE-MRI: the estimation of the tissue’s T1-time and local B1-field strength, and the estimation of the time-dependent concentration of contrast agent in the blood supply to the tissue of interest. In quantitative pharmacokinetic analysis, the perfusion and vascularization of tissues are estimated by measuring the response to an intravenous injection of contrast agent. This analysis relies on knowledge of the concentrations of contrast agent in both the tissue and in the blood perfusing the tissue. The contrast agent affects the T1 relaxation time of the tissue, and if the T1-time of a tissue is known, the concentration profile can be computed. However, local B1-inhomogeneities can affect the MRI signal strength, complicating the measurement of T1 using conventional methods. Furthermore, the inflow of fresh blood into the field of view causes an additional, location dependent signal enhancement in the blood, which makes a direct measurement of the T1-time (and thus the concentration) in blood impossible. This thesis introduces a new method to estimate a T1-map of tissues in the presence of B1-inhomogeneities. We do this by combining two MRI scans that can each be acquired within breath-holds: one that yields a precise T1-map, though biased by the inhomogeneous B1-field; and one that delivers an unbiased, but imprecise estimate. Combining the information of these two scans yields an estimate of the B1-field, which is then used to correct the T1-map. We validate our method in a phantom study, and in an in vivo study. We found that the proposed method successfully merges the high resolution of the first method with the insensitivity to B1-inhomogeneities of the second. This thesis also introduces a new method to estimate the time-dependent concentration of contrast agent in blood (i.e., the arterial input function (AIF)), which is affected by signal enhancement due to the inflow effect. We do this by first estimating the number of RF-pulses by incorporating knowledge about the average AIF in a population. We then use the number of pulses to re-estimate the concentration from the measured MRI signal, thereby correcting for the inflow effect. We validate our method by means of Monte Carlo simulations and with a controlled flow phantom experiment. We then apply our method to two patient datasets, and use the estimated arterial input function for pharmacokinetic modelling. The first dataset consisted of patients with spine related injuries, and was acquired under a variety of scan settings to assess the method’s robustness. The second dataset consisted of patients with Crohn’s Disease which had a clinically relevant CDEIS score available. In both datasets, we found that our method yields realistic pharmacokinetic model parameters. Instead, estimating the AIF from a distally placed region of interest, as is often done in literature, led to large variation and unrealistic parameters. Furthermore, in the Crohn’s patients we found a better correlation between the estimated pharmacokinetic parameter Ktrans and the CDEIS score, compared to traditional methods. Though the rationale for developing these methods were the presence of B1- inhomogeneities, and pronounced inflow effects in the aorta, other applications of pharmacokinetic modelling (e.g., in other parts of the body) may benefit from our methods, since they are generally applicable.@en