Viable tumor tissue detection in murine metastatic breast cancer by whole‐body MRI and multispectral analysis

Whole‐body MRI combined with a semiautomated hierarchical multispectral image analysis technique was evaluated as a method for detecting viable tumor tissue in a murine model of metastatic breast cancer (4T1 cell line). Whole‐body apparent diffusion coefficient, T₂, and proton density maps were acquired in this study. The viable tumor tissue segmentation included three‐stage k‐means clustering of the parametric maps, morphologic operations, application of a size threshold, and reader discrimination of the segmented objects. The segmentation results were validated by histologic evaluation, and the detection accuracy of the technique was evaluated at three size thresholds (15, 100, and 500 voxels). The accuracy was 88.9% for a 500‐voxel size threshold, and the area under receiver operating characteristic curve was 0.84. The regions of segmented viable tumor tissue within the primary tumors were found mostly on the periphery of the tumors in agreement with the histologic findings. The presented technique was found capable of detecting metastases and segmenting the viable tumor from necrotic regions within tumors found in this model. It offers a noninvasive, whole‐body, viable tumor tissue detection method for preclinical and potentially clinical applications such as tumor screening and evaluating therapeutic efficacy. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Assessment of the effects of cellular tissue properties on ADC measurements by numerical simulation of water diffusion

The apparent diffusion coefficient (ADC), as measured by diffusion‐weighted MRI, has proven useful in the diagnosis and evaluation of ischemic stroke. The ADC of tissue water is reduced by 30‐50% following ischemia and provides excellent contrast between normal and affected tissue. Despite its clinical utility, there is no consensus on the biophysical mechanism underlying the reduction in ADC. In this work, a numerical simulation of water diffusion is used to predict the effects of cellular tissue properties on experimentally measured ADC. The model indicates that the biophysical mechanisms responsible for changes in ADC postischemia depend upon the time over which diffusion is measured. At short diffusion times, the ADC is dependent upon the intrinsic intracellular diffusivity, while at longer, clinically relevant diffusion times, the ADC is highly dependent upon the cell volume fraction. The model also predicts that at clinically relevant diffusion times, the 30‐50% drop in ADC after ischemia can be accounted for by cell swelling alone when intracellular T₂ is allowed to be shorter than extracellular T₂. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Simultaneous estimation of T(2) and apparent diffusion coefficient in human articular cartilage in vivo with a modified three-dimensional double echo steady state (DESS) sequence at 3 T

T₂ mapping and diffusion‐weighted imaging complement morphological imaging for assessing cartilage disease and injury. The double echo steady state sequence has been used for morphological imaging and generates two echoes with markedly different T₂ and diffusion weighting. Modifying the spoiler gradient area and flip angle of the double echo steady state sequence allows greater control of the diffusion weighting of both echoes. Data from two acquisitions with different spoiler gradient areas and flip angles are used to simultaneously estimate the T₂ and apparent diffusion coefficient of each voxel. This method is verified in phantoms and validated in vivo in the knee; estimates from different regions of interest in the phantoms and cartilage are compared to those obtained using standard spin‐echo methods. The Pearson correlations were 0.984 for T₂ (～2% relative difference between spin‐echo and double echo steady state estimates) and 0.997 for apparent diffusion coefficient (?1% relative difference between spin‐echo and double echo steady state estimates) for the phantom study and 0.989 for T₂ and 0.987 for apparent diffusion coefficient in regions of interest in the human knee in vivo. High accuracy for simultaneous three‐dimensional T₂ and apparent diffusion coefficient measurements are demonstrated, while also providing morphologic three‐dimensional images without blurring or distortion in reasonable scan times. Magn Reson Med, 2011. © 2011 Wiley‐Liss, Inc.     

Aldehyde fixative solutions alter the water relaxation and diffusion properties of nervous tissue

Chemically‐fixed nervous tissues are well‐suited for high‐resolution, time‐intensive MRI acquisitions without motion artifacts, such as those required for brain atlas projects, but the aldehyde fixatives used may significantly alter tissue MRI properties. To test this hypothesis, this study characterized the impact of common aldehyde fixatives on the MRI properties of a rat brain slice model. Rat cortical slices immersion‐fixed in 4% formaldehyde demonstrated 21% and 81% reductions in tissue T₁ and T₂, respectively (P < 0.001). The T₂ reduction was reversed by washing slices with phosphate‐buffered saline (PBS) for 12 h to remove free formaldehyde solution. Diffusion MRI of cortical slices analyzed with a two‐compartment analytical model of water diffusion demonstrated 88% and 30% increases in extracellular apparent diffusion coefficient (ADC_EX) and apparent restriction size, respectively, when slices were chemically‐fixed with 4% formaldehyde (P ≤ 0.021). Further, fixation with 4% formaldehyde increased the transmembrane water exchange rate 239% (P < 0.001), indicating increased membrane permeability. Karnovsky's and 4% glutaraldehyde fixative solutions also changed the MRI properties of cortical slices, but significant differences were noted between the different fixative treatments (P < 0.05). The observed water relaxation and diffusion changes help better define the validity and limitations of using chemically‐fixed nervous tissue for MRI investigations. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Multicomponent T2 relaxation analysis in cartilage

MR techniques are sensitive to the early stages of osteoarthritis, characterized by disruption of collagen and loss of proteoglycan (PG), but are of limited specificity. Here, water compartments in normal and trypsin‐degraded bovine nasal cartilage were identified using a nonnegative least squares multiexponential analysis of T₂ relaxation. Three components were detected: T_2,1 = 2.3 ms, T_2,2 = 25.2 ms, and T_2,3 = 96.3 ms, with fractions w₁ = 6.2%, w₂ = 14.5%, and w₃ = 79.3%, respectively. Trypsinization resulted in increased (P < 0.01) values of T_2,2 = 64.2 ms and T_2,3 = 149.4 ms, supporting their assignment to water compartments that are bound and loosely associated with PG, respectively. The T₂ of the rapidly relaxing component was not altered by digestion, supporting assignment to relatively immobile collagen‐bound water. Relaxation data were simulated for a range of TE, number of echoes, and SNR to guide selection of acquisition parameters and assess the accuracy and precision of experimental results. Based on this, the expected experimental accuracy of measured T₂s and associated weights was within 2% and 4% respectively, with precision within 1% and 3%. These results demonstrate the potential of multiexponential T₂ analysis to increase the specificity of MR characterization of cartilage. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Multicomponent T2* mapping of knee cartilage: Technical feasibility ex vivo

Disorganization of collagen fibers is a sign of early‐stage cartilage degeneration in osteoarthritic knees. Water molecules trapped within well‐organized collagen fibrils would be sensitive to collagen alterations. Multicomponent effective transverse relaxation (T₂*) mapping with ultrashort echo time acquisitions is here proposed to probe short T₂ relaxations in those trapped water molecules. Six human tibial plateau explants were scanned on a 3T MRI scanner using a home‐developed ultrashort echo time sequence with echo times optimized via Monte Carlo simulations. Time constants and component intensities of T₂* decays were calculated at individual pixels, using the nonnegative least squares algorithm. Four T₂*‐decay types were found: 99% of cartilage pixels having mono‐, bi‐, or nonexponential decay, and 1% showing triexponential decay. Short T₂* was mainly in 1‐6 ms, while long T₂* was ～22 ms. A map of decay types presented spatial distribution of these T₂* decays. These results showed the technical feasibility of multicomponent T₂* mapping on human knee cartilage explants. Magn Reson Med, 2010. © 2010 Wiley‐Liss, Inc.     

Spatial distribution and relationship of T1ρ and T2 relaxation times in knee cartilage with osteoarthritis

T_1ρ and T₂ relaxation time constants have been proposed to probe biochemical changes in osteoarthritic cartilage. This study aimed to evaluate the spatial correlation and distribution of T_1ρ and T₂ values in osteoarthritic cartilage. Ten patients with osteoarthritis (OA) and 10 controls were studied at 3T. The spatial correlation of T_1ρ and T₂ values was investigated using Z‐scores. The spatial variation of T_1ρ and T₂ values in patellar cartilage was studied in different cartilage layers. The distribution of these relaxation time constants was measured using texture analysis parameters based on gray‐level co‐occurrence matrices (GLCM). The mean Z‐scores for T_1ρ and T₂ values were significantly higher in OA patients vs. controls (P < 0.05). Regional correlation coefficients of T_1ρ and T₂ Z‐scores showed a large range in both controls and OA patients (0.2–0.7). OA patients had significantly greater GLCM contrast and entropy of T_1ρ values than controls (P < 0.05). In summary, T_1ρ and T₂ values are not only increased but are also more heterogeneous in osteoarthritic cartilage. T_1ρ and T₂ values show different spatial distributions and may provide complementary information regarding cartilage degeneration in OA. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Model‐based blind estimation of kinetic parameters in dynamic contrast enhanced (DCE)‐MRI

A method to simultaneously estimate the arterial input function (AIF) and pharmacokinetic model parameters from dynamic contrast‐enhanced (DCE)‐MRI data was developed. This algorithm uses a parameterized functional form to model the AIF and k‐means clustering to classify tissue time‐concentration measurements into a set of characteristic curves. An iterative blind estimation algorithm alternately estimated parameters for the input function and the pharmacokinetic model. Computer simulations were used to investigate the algorithm's sensitivity to noise and initial estimates. In 12 patients with sarcomas, pharmacokinetic parameter estimates were compared with “truth” obtained from model regression using a measured AIF. When arterial voxels were included in the blind estimation algorithm, the resulting AIF was similar to the measured input function. The “true” K^trans values in tumor regions were not significantly different than the estimated values, 0.99 ± 0.41 and 0.86 ± 0.40 min^?1, respectively, P = 0.27. “True” k_ep values also matched closely, 0.70 ± 0.24 and 0.65 ± 0.25 min^?1, P = 0.08. When only tissue curves free of significant vascular contribution are used (v_p < 0.05), the resulting AIF showed substantial delay and dispersion consistent with a more local AIF such as has been observed in dynamic susceptibility contrast imaging in the brain. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Gd-DTPA2− as a measure of cartilage degradation

Glycosaminoglycans (GAGs) are the main source of tissue fixed charge density (FCD) in cartilage, and are lost early in arthritic diseases. We tested the hypothesis that, like Na⁺, the charged contrast agent Gd-DTPA^2- (and hence proton T₁) could be used to measure tissue FCD and hence GAG concentration. NMR spectroscopy studies of cartilage explants demonstrated that there was a strong correlation (r > 0.96) between proton T₁ in the presence of Gd-DTPA^2- and tissue sodium and GAG concentrations. An ideal one-compartment electrochemical (Donnan) equilibrium model was examined as a means of quantifying FCD from Gd-DTPA^2- concentration, yielding a value 50% less but linearly correlated with the validated method of quantifying FCD from Na⁺. These data could be used as the basis of an empirical model with which to quantify FCD from Gd-DTPA^2- concentration, or a more sophisticated physical model could be developed. Spatial distributions of FCD were easily observed in T₁-weighted MRI studies of trypsin and interleukin-1 induced cartilage degradation, with good histological correlation. Therefore, equilibration of the tissue in Gd-DTPA^2- gives us the opportunity to directly image (through T₁, weighting) the concentration of GAG, a major and critically important macromolecule in cartilage. Pilot clinical studies demonstrated Gd-DTPA^2- penetration into cartilage, suggesting that this technique is clinically feasible.     

Characterization of T2* heterogeneity in human brain white matter

Recent in vivo MRI studies at 7.0 T have demonstrated extensive heterogeneity of T₂* relaxation in white matter of the human brain. In order to study the origin of this heterogeneity, we performed T₂* measurements at 1.5, 3.0, and 7.0 T in normal volunteers. Formalin‐fixed brain tissue specimens were also studied using T₂*‐weighted MRI, histologic staining, chemical analysis, and electron microscopy. We found that T₂* relaxation rate (R₂* = 1/T₂*) in white matter in living human brain is linearly dependent on the main magnetic field strength, and the T₂* heterogeneity in white matter observed at 7.0 T can also be detected, albeit more weakly, at 1.5 and 3.0 T. The T₂* heterogeneity exists also in white matter of the formalin‐fixed brain tissue specimens, with prominent differences between the major fiber bundles such as the cingulum (CG) and the superior corona radiata. The white matter specimen with substantial difference in T₂* has no significant difference in the total iron content, as determined by chemical analysis. On the other hand, evidence from histologic staining and electron microscopy demonstrates these tissue specimens have apparent difference in myelin content and microstructure. Magn Reson Med, 2009. © 2009 Wiley‐Liss, Inc.     

Reduction of residual dipolar interaction in cartilage by spin‐lock technique

The influence of radiofrequency (RF) spin‐lock pulse on the laminar appearance of articular cartilage in MR images was investigated. Spin‐lock MRI experiments were performed on bovine cartilage plugs on a 4.7 Tesla small‐bore MRI scanner, and on human knee cartilage in vivo on a 1.5 Tesla clinical scanner. When the normal to the surface of cartilage was parallel to B₀, a typical laminar appearence was exhibited in T₂‐weighted images of cartilage plugs, but was absent in T_1ρ‐weighted images of the same plugs. At the “magic angle” orientation (when the normal to the surface of cartilage was 54.7° with respect to B₀), neither the T₂ nor the T_1ρ images demonstrated laminae. At the same time, T_1ρ values were greater than T₂ at both orientations throughout the cartilage. T_1ρ dispersion (i.e., the dependence of the relaxation rate on the spin‐lock frequency ω₁) was observed, which reached a steady‐state value of close to 2 kHz in both parallel and magic‐angle orientations. These results suggest that residual dipolar interaction from motionally‐restricted water and relaxation processes, such as chemical exchange, contribute to T_1ρ dispersion in cartilage. Further, one can reduce the laminar appearance in human articular cartilage by applying spin‐lock RF pulses, which may lead to a more accurate diagnosis of degenerative changes in cartilage. Magn Reson Med 52:1103–1109, 2004. © 2004 Wiley‐Liss, Inc.     

Fully automated reconstruction of ungated ghost magnetic resonance angiograms

Purpose:

To completely automate the reconstruction process during noncardiac‐gated unenhanced ghost magnetic resonance angiography (MRA).

Materials and Methods:

Ungated unenhanced ghost MRA of the calf was performed in 16 volunteers. K‐means and fuzzy c‐means (FCM) clustering algorithms using prominent image features were applied to automatically create angiograms of the calf in volunteers undergoing ungated ghost MRA. Ghost angiograms reconstructed automatically were compared to those created manually on the basis of diagnostic image quality and apparent arterial‐to‐background contrast‐to‐noise ratio (CNR). Images were also ranked by an expert user in their order of preference using an ordinal scale.

Results:

Compared with the ghost angiograms created manually, ghost angiograms reconstructed automatically with the use of clustering analysis provided similar arterial‐to‐background CNR values. No differences in diagnostic quality or preference were identified between images reconstructed manually and automatically.

Conclusion:

We present fully automated image reconstruction algorithms for use with ungated and unenhanced ghost MRA. These automated algorithms, based on the use of k‐means or FCM clustering, can be used to eliminate manual postprocessing that is time‐consuming and subject to variability. J. Magn. Reson. Imaging 2010;31:655–662. © 2010 Wiley‐Liss, Inc.     

Further studies on the anisotropic distribution of collagen in articular cartilage by μMRI

To further study the anisotropic distribution of the collagen matrix in articular cartilage, microscopic magnetic resonance imaging experiments were carried out on articular cartilages from the central load‐bearing area of three canine humeral heads at 13 μm resolution across the depth of tissue. Quantitative T₂ images were acquired when the tissue blocks were rotated, relative to B₀, along two orthogonal directions, both perpendicular to the normal axis of the articular surface. The T₂ relaxation rate (R₂) was modeled, by three fibril structural configurations (solid cone, funnel, and fan), to represent the anisotropy of the collagen fibrils in cartilage from the articular surface to the cartilage/bone interface. A set of complex and depth‐dependent characteristics of collagen distribution was found in articular cartilage. In particular, there were two anisotropic components in the superficial zone and an asymmetrical component in the radial zone of cartilage. A complex model of the three‐dimensional fibril architecture in articular cartilage is proposed, which has a leaf‐like or layer‐like structure in the radial zone, arises in a radial manner from the subchondral bone, spreads and arches passing the isotropic transitional zone, and exhibits two distinct anisotropic components (vertical and transverse) in the surface portion of the tissue. Magn Reson Med, 2011. © 2010 Wiley‐Liss, Inc.     

Fast diffusion‐weighted steady state free precession imaging of in vivo knee cartilage

Quantification of molecular diffusion with steady state free precession (SSFP) is complicated by the fact that diffusion effects accumulate over several repetition times (TR) leading to complex signal dependencies on transverse and longitudinal magnetization paths. This issue is commonly addressed by setting TR > T₂, yielding strong attenuation of all higher modes, except of the shortest ones. As a result, signal attenuation from diffusion becomes T₂ independent but signal‐to‐noise ratio (SNR) and sequence efficiency are remarkably poor. In this work, we present a new approach for fast in vivo steady state free precession diffusion‐weighted imaging of cartilage with TR << T₂ offering a considerable increase in signal‐to‐noise ratio and sequence efficiency. At a first glance, prominent coupling between magnetization paths seems to complicate quantification issues in this limit, however, it is observed that diffusion effects become rather T₂(ΔD ～ 1/10 ΔT₂) but not T₁ independent (ΔD ～ 1/2 ΔT₁) for low flip angles α ～ 10 ? 15°. As a result, fast high‐resolution (0.35 × 0.35 ? 0.50 × 0.50 mm² in‐plane resolution) quantitative diffusion‐weighted imaging of human articular cartilage is demonstrated at 3.0 T in a clinical setup using estimated T₁ and T₂ or a combination of measured T₁ and estimated T₂ values. Magn Reson Med, 2012. © 2011 Wiley Periodicals, Inc.     

Longitudinal analysis of MRI T2 knee cartilage laminar organization in a subset of patients from the osteoarthritis initiative: A texture approach

Cartilage magnetic resonance imaging T₂ relaxation time is sensitive to hydration, collagen content, and tissue anisotropy, and a potential imaging‐based biomarker for knee osteoarthritis. This longitudinal pilot study presents an improved cartilage flattening technique that facilitates texture analysis using gray‐level co‐occurrence matrices parallel and perpendicular to the cartilage layers, and the application of this technique to the knee cartilage of 13 subjects of the osteoarthritis initiative at baseline, 1‐year follow‐up, and 2‐year follow‐up. Cartilage flattening showed minimum distortion (～ 0.5 ms) of mean T₂ values between nonflattened and flattened T₂ maps. Gray‐level co‐occurrence matrices texture analysis of flattened T₂ maps detected a cartilage laminar organization at baseline, 1‐year follow‐up, and 2‐year follow‐up by yielding significant (P < 0.05) differences between texture parameters perpendicular and parallel to the cartilage layers. Tendencies showed higher contrast, dissimilarity, angular second moment, and energy perpendicular to the cartilage layers; and higher homogeneity, entropy, variance, and correlation parallel to them. Significant (P < 0.05) longitudinal texture changes were also detected reflecting subtle signs of a laminar disruption. Tendencies showed decreasing contrast, dissimilarity, and entropy; and increasing homogeneity, energy, and correlation. Results of this study warrant further investigation to complete the assessment of the usefulness of the presented methodology in the study of knee osteoarthritis. Magn Reson Med, 2010. © 2010 Wiley‐Liss, Inc.     

Strain‐dependent T1 relaxation profiles in articular cartilage by MRI at microscopic resolutions

To investigate the dependency of T₁ relaxation on mechanical strain in articular cartilage, quantitative magnetic resonance T₁ imaging experiments were carried out on cartilage before/after the tissue was immersed in gadolinium contrast agent and when the tissue was being compressed (up to ～48% strains). The spatial resolution across the cartilage depth was 17.6 μm. The T₁ profile in native tissue (without the presence of gadolinium ions) was strongly strain‐dependent, which is also depth‐dependent. At the modest strains (e.g., 14% strain), T₁ reduced by up to 68% in the most surface portion of the tissue. Further compression (e.g., 45% strain) reduced T₁ mostly in the middle and deep portions of the tissue. For the gadolinium‐immersed tissue, both modest and heavy compressions (up to 48% strain) increased T₁ slightly but significantly, although the overall shapes of the T₁ profiles remained approximately the same regardless of the amount of strains. The complex relationships between the T₁ profiles and the mechanical strains were a direct consequence of the depth‐dependent proteoglycan concentration in the tissue, which determined the tissue's mechanical properties. This finding has potential implications in the use of gadolinium contrast agent in clinical magnetic resonance imaging of cartilage (the dGEMRIC procedure), when the loading or loading history of patients is considered. Magn Reson Med, 2011. © 2011 Wiley‐Liss, Inc.     

T1ρ MRI of human musculoskeletal system

Magnetic resonance imaging (MRI) offers the direct visualization of the human musculoskeletal (MSK) system, especially all diarthrodial tissues including cartilage, bone, menisci, ligaments, tendon, hip, synovium, etc. Conventional MRI techniques based on T₁‐ and T₂‐weighted, proton density (PD) contrast are inconclusive in quantifying early biochemically degenerative changes in MSK system in general and articular cartilage in particular. In recent years, quantitative MR parameter mapping techniques have been used to quantify the biochemical changes in articular cartilage, with a special emphasis on evaluating joint injury, cartilage degeneration, and soft tissue repair. In this article we focus on cartilage biochemical composition, basic principles of T_1ρ MRI, implementation of T_1ρ pulse sequences, biochemical validation, and summarize the potential applications of the T_1ρ MRI technique in MSK diseases including osteoarthritis (OA), anterior cruciate ligament (ACL) injury, and knee joint repair. Finally, we also review the potential advantages, challenges, and future prospects of T_1ρ MRI for widespread clinical translation. J. Magn. Reson. Imaging 2015;41:586–600. © 2014 Wiley Periodicals, Inc.     

T2 measurement in articular cartilage: Impact of the fitting method on accuracy and precision at low SNR

T₂ relaxation time is a promising MRI parameter for the detection of cartilage degeneration in osteoarthritis. However, the accuracy and precision of the measured T₂ may be substantially impaired by the low signal‐to‐noise ratio of images available from clinical examinations. The purpose of this work was to assess the accuracy and precision of the traditional fit methods (linear least‐squares regression and nonlinear fit to an exponential) and two new noise‐corrected fit methods: fit to a noise‐corrected exponential and fit of the noise‐corrected squared signal intensity to an exponential. Accuracy and precision have been analyzed in simulations, in phantom measurements, and in seven repetitive acquisitions of the patellar cartilage in six healthy volunteers. Traditional fit methods lead to a poor accuracy for low T₂, with overestimations of the exact T₂ up to 500%. The noise‐corrected fit methods demonstrate a very good accuracy for all T₂ values and signal‐to‐noise ratio. Even more, the fit to a noise‐corrected exponential results in precisions comparable to the best achievable precisions (Cramér‐Rao lower bound). For in vivo images, the traditional fit methods considerably overestimate T₂ near the bone‐cartilage interface. Therefore, using an adequate fit method may substantially improve the sensitivity of T₂ to detect pathology in cartilage and change in T₂ follow‐up examinations. Magn Reson Med, 2010. © 2009 Wiley‐Liss, Inc.     

T1ρ-relaxation in articular cartilage: Effects of enzymatic degradation

Spin-lattice relaxation in the rotating frame (T_1ρ) dispersion spectroscopy and imaging were used to study normal and enzymatically degraded bovine articular cartilage. Normal specimens demonstrate significant T_1ρ “dispersion” (～60 to ～130 ms) in the 100 Hz to 9 kHz frequency range. Proteoglycan-degraded specimens have 33% greater T_1ρ values than collagen-degraded or normal samples. T_1ρ-weighted images reveal structure not found in conventional T₁-or T₂-weighted images. Our results suggest that T_1ρ measurements are selectively sensitive to proteoglycan content. The potential of this method in distinguishing the early degenerative changes in cartilage associated with osteoarthritis is discussed.     

Human whole blood T2 relaxometry at 3 Tesla

A precise understanding of human blood spin–spin relaxation is of major importance for numerous applications, particularly functional magnetic resonance imaging (fMRI), which is increasingly performed at 3 Tesla. It is well known that T₂ measured from partially deoxygenated blood depends on the Carr–Purcell Meiboom–Gill (CPMG) refocusing interval (τ₁₈₀) and on blood oxygenation (Y), yet debate remains over the quantification of this phenomenon, primarily with respect to the accuracy of its characterization by the diffusion and fast two‐site exchange models. In this study, a detailed characterization of the deoxygenation‐induced T₂ reduction in human whole blood, as well as a comprehensive assessment of the role of τ₁₈₀, were performed at 3 T. The diffusion model was found to better fit the observed T₂ behavior as compared with the exchange model. The estimated diffusion‐model parameters suggest the T₂ decay enhancement at 3 T is due to a linear increase in the magnitude of deoxygenation‐induced field inhomogeneities with field strength. These findings also confirm the potential of τ₁₈₀ manipulation in measuring changes in venous blood volume. Magn Reson Med 61:249–254, 2009. © 2009 Wiley‐Liss, Inc.