Determination of arterial input function using fuzzy clustering for quantification of cerebral blood flow with dynamic susceptibility contrast-enhanced MR imaging

An accurate determination of the arterial input function (AIF) is necessary for quantification of cerebral blood flow (CBF) using dynamic susceptibility contrast-enhanced magnetic resonance imaging. In this study, we developed a method for obtaining the AIF automatically using fuzzy c-means (FCM) clustering. The validity of this approach was investigated with computer simulations. We found that this method can automatically extract the AIF, even under very noisy conditions, e.g., when the signal-to-noise ratio is 2. The simulation results also indicated that when using a manual drawing of a region of interest (ROI) (manual ROI method), the contamination of surrounding pixels (background) into ROI caused considerable overestimation of CBF. We applied this method to six subjects and compared it with the manual ROI method. The CBF values, calculated using the AIF obtained using the manual ROI method [CBF(manual)], were significantly higher than those obtained with FCM clustering [CBF(fuzzy)]. This may have been due to the contamination of non-arterial pixels into the manually drawn ROI, as suggested by simulation results. The ratio of CBF(manual) to CBF(fuzzy) ranged from 0.99-1.83 [1.31 +/- 0.26 (mean +/- SD)]. In conclusion, our FCM clustering method appears promising for determination of AIF because it allows automatic, rapid and accurate extraction of arterial pixels. J. Magn. Reson. Imaging 2001;13:797-806.     

Correction of arterial input function in dynamic contrast-enhanced MRI of the liver

Purpose:

To develop a postprocessing method to correct saturation of arterial input function (AIF) in T1‐weighted dynamic contrast‐enhanced magnetic resonance imaging (DCE‐MRI) for quantification of hepatic perfusion.

Materials and Methods:

The saturated AIF is corrected by parameterizing the first pass of the AIF as a smooth function with a single peak and minimizing a least‐squares error in fitting the liver DCE‐MRI data to a dual‐input single‐compartment model. Sensitivities of the method to the degree of saturation in the AIF first‐pass peak and the image contrast‐to‐noise ratio were assessed. The method was also evaluated by correlating portal venous perfusion with an independent overall liver function measurement.

Results:

The proposed method corrects the distorted AIF with a saturation ratio up to 0.45. The corrected AIF improved hepatic arterial perfusion by ?23.4% and portal venous perfusion by 26.9% in a study of 12 patients with liver cancers. The correlation between the mean voxelwise portal venous perfusion and overall liver function measurement was improved by using the corrected AIFs (R² = 0.67) compared with the saturated AIFs (R² = 0.39).

Conclusion:

The method is robust for correcting AIF distortion and has the potential to improve quantification of hepatic perfusion for assessment of liver tissue response to treatment in patients with hepatic cancers. J. Magn. Reson. Imaging 2012;36:411–421. © 2012 Wiley Periodicals, Inc.

     

Impact of the arterial input function on microvascularization parameter measurements using dynamic contrast-enhanced ultrasonography

AIM: To evaluate the sources of variation influencing the microvascularization parameters measured by dynamic contrast-enhanced ultrasonography (DCE-US). METHODS: Firstly, we evaluated, in vitro , the impact of the manual repositioning of the ultrasound probe and the variations in flow rates. Experiments were conducted using a custom-made phantom setup simulating a tumor and its associated arterial input. Secondly, we evaluated, in vivo , the impact of multiple contrast agent injections and of examination day, as well as the influence of the size of region of interest (ROI) associated with the arterial input function (AIF). Experiments were conducted on xenografted B16F10 female nude mice. For all of the experiments, an ultrasound scanner along with a linear transducer was used to perform pulse inversion imaging based on linear raw data throughout the experiments. Semi-quantitative and quantitative analyses were performed using two signal-processing methods. RESULTS:In vitro , no microvascularization parameters, whether semi-quantitative or quantitative, were significantly correlated (P values from 0.059 to 0.860) with the repositioning of the probe. In addition, all semiquantitative microvascularization parameters were correlated with the flow variation while only one quantitative parameter, the tumor blood flow, exhibited P value lower than 0.05 (P = 0.004). In vivo , multiple contrast agent injections had no significant impact (P values from 0.060 to 0.885) on microvascularization parameters. In addition, it was demonstrated that semi-quantitative microvascularization parameters were correlated with the tumor growth while among the quantitative parameters, only the tissue blood flow exhibited P value lower than 0.05 (P = 0.015). Based on these results, it was demonstrated that the ROI size of the AIF had significant influence on microvascularization parameters: in the context of larger arterial ROI (from 1.17 ± 0.6 mm 3 to 3.65 ± 0.3 mm 3 ), tumor blood flow and tumor blood volume were correlated with the tumor growth, exhibiting P values lower than 0.001. CONCLUSION: AIF selection is an essential aspect of the deconvolution process to validate the quantitative DCE-US method.     

Misregistration artifacts in image-derived arterial input function in non-echo-planar imaging-based dynamic contrast-enhanced MRI

PURPOSE: To characterize misregistration artifact in arterial input function (AIF) pixels in dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) using a two-dimensional non-echo-planar imaging (EPI)-based gradient-recalled echo (GRE) sequence. MATERIALS AND METHODS: Dynamic gadopentetate-enhanced MRI was acquired in the rat using a semikeyhole acquisition scheme. The AIF was obtained from abdominal aorta pixels. Different sliding-window reconstruction techniques were applied to determine which lines in a series of the semikeyhole acquisition were associated with the misregistration artifacts. RESULTS: The misregistration along the phase-encoding direction arose when k-space lines were acquired during the rise-time of the aortic gadolinium concentration. The maximum blood concentration of gadolinium estimated from the phase shift calculation agreed with that estimated from dosage. CONCLUSION: AIF misregistration results from a phase shift due to increasing gadolinium concentration in the aorta, and may need to be considered in small animal DCE-MRI studies with a high rate of rise in the AIF in high-field MR applications.     

3．0TMR动脉自旋标记与动态磁敏感对比增强灌注技术在脑胶质瘤术前分级中的对照研究

目的：对照研究动脉自旋标记（arterial spin labeling,ASL）与动态磁敏感对比增强（dynamic susceptibility contrast—enhanced．DSC）灌注成像技术在脑胶质瘤中的灌注特点．探讨ASL在脑胶质瘤术前分级中的临床应用价值。方法：使用3．0TMR成像系统对23例脑胶质瘤患者（术后病理证实高级别胶质瘤17例,低级别胶质瘤6例）术前行常规扫描外,加扫ASL及DSC灌注检查。测量肿瘤实质部分最大肿瘤血流量（maximal tumor blood flow,TBFmax）以及对侧白质、对侧灰质、对侧半球的血流量（cerebral blood flow,CBF）。结果：23例脑胶质瘤患者。两种灌注方法均获得了一致的灌注结果,TBF max／对侧白质CBF、TBFmax／对侧灰质CBF及TBFmax／对侧半球CBF的各比值在ASL和DSC两种技术之间的差异无明显统计学意义（P〉0．05）,但在高、低级别胶质瘤之间的差异均有统计学意义（P〈0．05）。在ASL法中,TBFmax／对侧白质CBF、TBFmax／对侧灰质CBF及TBFmax／对侧半球CBF分别取阈值为3．06、0．46和1．31时,其敏感性分别为i00％、88．2％和100％．特异性分别为83．3％、83．3％和100％。结论：ASL在评估脑胶质瘤血流灌注方面与DSC之间有相似的敏感性,具有可重复性高、完全无创性等优点,同时有助于术前对脑胶质瘤进行分级评判。     

Estimating the arterial input function using two reference tissues in dynamic contrast-enhanced MRI studies: fundamental concepts and simulations.

In dynamic contrast-enhanced MRI (DCE-MRI) studies, an accurate knowledge of the arterial contrast agent concentration as a function of time is crucial for the estimation of kinetic parameters. In this work, a novel method for estimating the arterial input function (AIF) based on the contrast agent concentration-vs.-time curves in two different reference tissues is described. It is assumed that the AIFs of the two tissues have the same shape, and that simple models with two or more compartments, and unknown kinetic parameters, can describe their tracer concentration-vs.-time curves. Based on the principle of self-consistency, one can relate the two tracer concentration-vs.-time curves to estimate their common underlining AIF, together with the kinetic parameters of the two tissues. In practice, the measured concentration-vs.-time curves have noise, and the AIFs of the two tissues are not exactly the same due to different dispersion effects. These factors will produce errors in the AIF estimate. Simulation studies show that despite the two error sources, the double-reference-tissue method provides reliable estimates of the AIF.     

Phase‐based arterial input function measurements for dynamic susceptibility contrast MRI

In dynamic susceptibility contrast perfusion MRI, arterial input function (AIF) measurements using the phase of the MR signal are traditionally performed inside an artery. However, phase‐based AIF selection is also feasible in tissue surrounding an artery such as the middle cerebral artery, which runs approximately perpendicular to B₀ since contrast agents also induce local field changes in tissue surrounding the artery. The aim of this study was to investigate whether phase‐based AIF selection is better performed in tissue just outside the middle cerebral artery than inside the artery. Additionally, phase‐based AIF selection was compared to magnitude‐based AIF selection. Both issues were studied theoretically and using numerical simulations, producing results that were validated using phantom experiments. Finally, an in vivo experiment was performed to illustrate the feasibility of phase‐based AIF selection. Three main findings are presented: first, phase‐based AIF selections are better made in tissue outside the middle cerebral artery, rather than within the middle cerebral artery, since in the latter approach partial‐volume effects affect the shape of the estimated AIF. Second, optimal locations for phase‐based AIF selection are similar for different clinical dynamic susceptibility contrast MRI sequences. Third, phase‐based AIF selection allows more locations in tissue to be chosen that show the correct AIF than does magnitude‐based AIF selection. Magn Reson Med, 2010. © 2010 Wiley‐Liss, Inc.     

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.     

3.0T 磁共振动态增强扫描定量分析在胰腺癌中的应用

目的：探讨磁共振动态增强扫描定量参数在胰腺癌中的应用价值。方法用3.0T 磁共振对病理证实的27例胰腺癌患者进行动态增强扫描,图像通过 Jims 软件的 Toft with Vp 模型分别计算病灶和正常胰腺组织的定量参数：K trans ,k ep ,Ve ,Vp ,并应用 SPSS17.0软件进行单向方差分析。结果胰腺癌的 K trans 值、k ep 值、Ve 值、Vp 值分别为：(0.303±0.321)min,(1.387±1.486)min,(25.07±10.98)%和(3.420±4.692)%;而正常胰腺组织的 K trans 值、k ep 值、Ve 值、Vp 值分别为：(1.235±0.777)min,(9.277±7.996)min,(17.89±8.882)%,(7.196±6.704)%,胰腺癌及正常胰腺组织的各参数间均存在显著的统计学差异(F 值分别为33.188,25.414,6.984,5.78,P 值均<0.05)。结论胰腺癌磁共振动态增强扫描定量参数能够准确反映病灶血流灌注及微循环变化,有助于不典型病变的鉴别诊断。     

RF excitation profiles with FAIR: impact of truncation of the arterial input function on quantitative perfusion

This study investigates the impact of imaging coil length and consequent truncation of the arterial input function on the perfusion signal contrast obtained in the flow-sensitive alternating inversion recovery (FAIR) perfusion imaging measurement. We examined the difference in perfusion contrast achieved with head, head and neck, and body imaging coils based on the hypothesis that the standard head coil provides a truncated input function compared with that provided by the body coil and that this effect will be accentuated at long inversion times. The TI-dependent cerebral response of the FAIR sequence was examined at 1.5 T by varying the TI from 200 to 3500 msec with both the head and whole body coils (n = 5) as well as using a head and neck coil (n = 3). Difference signal intensity DeltaM and quantitative cerebral blood flow (CBF) were plotted against TI for each coil configuration. Despite a lower signal-to-noise ratio, relative CBF was significantly greater when measured with the body or head and neck coil compared with the standard head coil for longer inversion times (two-way ANOVA, P < or = 0.002). This effect is attributed to truncation of the arterial input function of labeled water by the standard head coil and the resultant inflow of unlabeled spins to the image slice during control image acquisition, resulting in overestimation of CBF. The results support the conclusion that the arterial input function depends on the anatomic extent of the inversion pulse in FAIR, particularly at longer mixing times (TI > 1200 msec at 1.5 T). Use of a head and neck coil ensures adequate inversion while preserving SNR that is lost in the body coil.     

Evaluation of signal formation in local arterial input function measurements of dynamic susceptibility contrast MRI

Correct arterial input function (AIF) measurements in dynamic susceptibility contrast‐MRI are crucial for quantification of the hemodynamic parameters. Often a single global AIF is selected near a large brain‐feeding artery. Alternatively, local AIF measurements aim for voxel‐specific AIFs from smaller arteries. Because local AIFs are measured higher in the arterial‐tree, it is assumed that these will reflect the true input of the microvasculature much better. However, do the measured local AIFs reflect the true concentration‐time curves of small arteries? To answer this question, in vivo data were used to evaluate local AIF candidates selected based on two different types of angiograms. For interpretation purposes, a 3D numerical model that simulated partial‐volume effects in local AIF measurements was created and the simulated local AIFs were compared to the ground truth. The findings are 2‐fold. First, the in vivo data showed that the shape‐characteristics of local AIFs are similar to the shape‐characteristics of gray matter concentration‐time curves. Second, these findings are supported by the simulations showing broadening of the measured local AIFs compared to the ground truth. These findings are suggesting that local AIF measurements do not necessarily reflect the true concentration‐time curve in small arteries. Magn Reson Med, 2012. © 2011 Wiley Periodicals, Inc.     

Feasibility of using limited-population-based arterial input function for pharmacokinetic modeling of osteosarcoma dynamic contrast-enhanced MRI data.

For clinical dynamic contrast-enhanced (DCE) MRI studies, it is often not possible to obtain reliable arterial input function (AIF) in each measurement. Thus, it is important to find a representative AIF for pharmacokinetic modeling of DCE-MRI data when individual AIF (Ind-AIF) measurements are not available. A total of 16 patients with osteosarcomas in the lower extremity (knee region) underwent multislice DCE-MRI. Reliable Ind-AIFs were obtained in five patients with a contrast injection rate of 2 cc/s and another five patients with a 1 cc/s injection rate. Average AIF (Avg-AIF) for each injection rate was constructed from the corresponding five Ind-AIFs. For each injection rate there are no statistically significant differences between pharmacokinetic parameters of the five patients derived with Ind-AIFs and Avg-AIF. There are no statistically significant changes in pharmacokinetic parameters of the 16 patients when the two Avg-AIFs were applied in kinetic modeling. The results suggest that it is feasible, as well as practical, to use a limited-population-based Avg-AIF for pharmacokinetic modeling of osteosarcoma DCE-MRI data. Further validation with a larger population and multiple regions is desirable.     

New criterion to aid manual and automatic selection of the arterial input function in dynamic susceptibility contrast MRI

Dynamic susceptibility contrast‐MRI requires an arterial input function (AIF) to obtain cerebral blood flow, cerebral blood volume, and mean transit time. The current AIF selection criteria discriminate venous, capillary, and arterial profiles based on shape and timing characteristics of the first passage. Unfortunately, partial volume effects can lead to shape errors in the bolus passage, including a narrower and higher peak, which might be selected as a “correct” AIF. In this study, a new criterion is proposed that detects shape errors based on tracer kinetic principles for computing cerebral blood volume. This criterion uses the ratio of the steady‐state value to the area‐under‐the‐curve of the first passage, which should result in an equal value for tissue and arterial responses. By using a reference value from tissue, partial volume effects–induced shape errors of the AIF measurement can be detected. Different factors affecting the ratio were investigated using simulations. These showed that the new criterion should only be used in studies with T₁‐insensitive acquisition. In vivo data were used to evaluate the proposed approach. The data showed that the new criterion enables detection of shape errors, although false positives do occur, which could be easily avoided when combined with current AIF selection criteria. Magn Reson Med, 2011. © 2010 Wiley‐Liss, Inc.     

Differentiation of low-grade oligodendrogliomas from low-grade astrocytomas by using quantitative blood-volume measurements derived from dynamic susceptibility contrast-enhanced MR imaging

BACKGROUND AND PURPOSE: Histopathologic evaluation remains the reference standard for diagnosis of glioma and classification of histologic subtypes, but is challenged by subjective criteria, tissue sampling error, and lack of specific tumor markers. Anatomic imaging is essential for surgical planning of gliomas but is limited by its nonspecificity and its inability to depict beyond morphologic aberrations. The purpose of our study was to investigate dynamic susceptibility contrast-enhanced (DSC) MR imaging characteristics of the two most common subtypes of low-grade infiltrating glioma: astrocytoma and oligodendroglioma. We hypothesized that tumor blood-volume measurements, derived from DSC MR imaging, would help differentiate the two on the basis of differences in tumor vascularity. METHODS: We studied 25 consecutive patients with treatment-naive, histopathologically confirmed World Health Organization grade II astrocytoma (n = 11) or oligodendroglioma (n = 14). All patients underwent anatomic and DSC MR imaging immediately before surgical resection. Histologic confirmation was obtained in all patients. Anatomic MR images were analyzed for morphologic features, and DSC MR data were processed to yield quantitative cerebral blood volume (CBV) measurements. RESULTS: The maximum relative CBV (rCBV(max)) in tumor ranged from 0.48 to 1.34 (0.92 +/- 0.27, median +/- SD) in astrocytomas and from 1.29 to 9.24 (3.68 +/- 2.39) in oligodendrogliomas. The difference in median rCBV(max) between the two tumor types was significant (P < .0001). CONCLUSION: The tumor rCBV(max) measurements derived from DSC MR imaging were significantly higher in low-grade oligodendrogliomas than in astrocytomas. Our findings suggest that tumor rCBV(max) derived from DSC MR imaging can be used to distinguish between the two low-grade gliomas.     

New subtraction algorithms for evaluation of lesions on dynamic contrast-enhanced MR mammography

We report new subtraction algorithms for the detection of lesions in dynamic contrast-enhanced MR mammography(CE MRM). Twenty-five patients with suspicious breast lesions underwent dynamic CE MRM using 3D fast low-angle shot. After the acquisition of the T1-weighted scout images, dynamic images were acquired six times after the bolus injection of contrast media. Serial subtractions, step-by-step subtractions, and reverse subtractions, were performed. Two radiologists attempted to differentiate benign from malignant lesion in consensus. The sensitivity, specificity, and accuracy of the method leading to the differentiation of malignant tumor from benign lesions were 85.7, 100, and 96%, respectively. Subtraction images allowed for better visualization of the enhancement as well as its temporal pattern than visual inspection of dynamic images alone. Our findings suggest that the new subtraction algorithm is adequate for screening malignant breast lesions and can potentially replace the time–intensity profile analysis on user-selected regions of interest. Electronic Publication     

Methods for quantitative assessment of tissue microcirculation using dynamic contrast-enhanced MR imaging

Summary The development of rapid magnetic resonance imaging (MRI) sequences makes it possible to detect the fast kinetics of tissue response after intraveneous administration of paramagnetic contrast media (CM), reflecting the status of tissue microcirculation. In this paper, the basic physical and tracer kinetic principles of dynamic relaxivity and susceptibility contrast MRI techniques are reviewed. The quantitative analysis of the acquired dynamic image data is broken up into an MR specific part, in which the observed signal variations are related to the CM concentration in the tissue, and an MR independent part, in which the computed concentration-time-courses are analyzed by tracer kinetic modeling. The purpose of the applied models is to describe the underlying physiological processes in mathematical terms and thus to enable the estimation of tissue specific parameters from measured dynamic image series. Whereas the capillary permeability can be estimated from dynamic relaxivity contrast enhanced MRI studies, the regional blood volume as well as the regional blood flow can be determined from dynamic susceptibility contrast enhanced image series. However, since there are no intravascular but only diffusible CM available at present, the application of the susceptibility technique is currently restricted to brain tissues with intact blood brain barrier. The practical realization of both dynamic MRI techniques is demonstrated by case studies. Eingegangen am 5. M?rz 1997 Angenommen am 24. April 1997     

Effects of artery input function on dynamic contrast-enhanced MRI for determining grades of gliomas

Objective:To evaluate the effect of artery input function (AIF) derived from different arteries for pharmacokinetic modeling on dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parameters in the grading of gliomas.Methods:49 patients with pathologically confirmed gliomas were recruited and underwent DCE-MRI. A modified Tofts model with different AIFs derived from anterior cerebral artery (ACA), ipsilateral and contralateral middle cerebral artery (MCA) and posterior cerebral artery (PCA) was used to estimate quantitative parameters such as K^trans (volume transfer constant) and V_e (fractional extracellular-extravascular space volume) for distinguishing the low grade glioma from high grade glioma. The K^trans and V_e were compared between different arteries using Two Related Samples Tests (TRST) (i.e. Wilcoxon Signed Ranks Test). In addition, these parameters were compared between the low and high grades as well as between the grade II and III using the Mann-Whitney U-test. A p-value of less than 0.05 was regarded as statistically significant.Results:All the patients completed the DCE-MRI successfully. Sharp wash-in and wash-out phases were observed in all AIFs derived from the different arteries. The quantitative parameters (K^trans and V_e) calculated from PCA were significant higher than those from ACA and MCA for low and high grades, respectively (p < 0.05). Despite the differences of quantitative parameters derived from ACA, MCA and PCA, the K^trans and V_e from any AIFs could distinguish between low and high grade, however, only K^trans from any AIFs could distinguish grades II and III. There was no significant correlation between parameters and the distance from the artery, which the AIF was extracted, to the tumor.Conclusion:Both quantitative parameters K^trans and V_e calculated using any AIF of ACA, MCA, and PCA can be used for distinguishing the low- from high-grade gliomas, however, only K^trans can distinguish grades II and III.Advances in knowledge:We sought to assess the effect of AIF on DCE-MRI for determining grades of gliomas. Both quantitative parameters K^trans and V_e calculated using any AIF of ACA, MCA, and PCA can be used for distinguishing the low- from high-grade gliomas.     

Iron-induced susceptibility effect at the globus pallidus causes underestimation of flow and volume on dynamic susceptibility contrast-enhanced MR perfusion images

BACKGROUND AND PURPOSE: Age-related iron accumulation in extrapyramidal nuclei causes T2 shortening, which may result in decreased signal intensity in these areas on MR images. Because the dynamic susceptibility contrast-enhanced technique uses heavily T2*- or T2-weighted images, the iron-induced susceptibility may have direct impact on perfusion imaging. The purpose of this study was to assess the effect of iron-induced susceptibility on the calculated perfusion parameters. The difference of this effect between gradient-echo and spin-echo sequences was also assessed. METHODS: Dynamic susceptibility contrast-enhanced MR perfusion imaging data of 12 patients were used for this study. Perfusion images were obtained using a single shot spin-echo echo-planar imaging sequence in seven patients and a gradient-echo echo-planar imaging sequence in five patients. Region of interest measurements of relative cerebral blood flow, relative cerebral blood volume, and mean transit time were obtained at various parts of the gray matter, including the globus pallidus, putamen, caudate nucleus, thalamus, and cerebral cortex of temporal, frontal, and occipital lobes. The baseline signal intensity on the source images and the magnitude of signal change (DeltaR2* or DeltaR2) were also assessed. RESULTS: The globus pallidus had statistically significantly lower values of relative cerebral blood flow, relative cerebral blood volume, baseline signal intensity, and magnitude of signal change compared with other parts of the gray matter for both gradient-echo and spin-echo sequences (P <.05). Underestimations of these values were more prominent for the gradient-echo than for the spin-echo sequence. Little variance in the measured mean transit time was noted. CONCLUSION: Iron-induced susceptibility effect may lead to underestimation of relative cerebral blood flow and relative cerebral blood volume in the basal ganglia.