Clinical application and technical considerations of T₁ & T₂(*) mapping in cardiac, liver, and renal imaging

Abstract

Pathological tissue alterations due to disease processes such as fibrosis, edema and infiltrative disease can be non-invasively visualized and quantified by MRI using T₁ and T₂ relaxation properties. Pixel-wise mapping of T₁ and T₂ image sequences enable direct quantification of T₁, T₂(*), and extracellular volume values of the target organ of interest. Tissue characterization based on T₁ and T₂(*) mapping is currently making the transition from a research tool to a clinical modality, as clinical usefulness has been established for several diseases such as myocarditis, amyloidosis, Anderson-Fabry and iron deposition. Other potential clinical applications besides the heart include, quantification of steatosis, cirrhosis, hepatic siderosis and renal fibrosis. Here, we provide an overview of potential clinical applications of T₁ andT₂(*) mapping for imaging of cardiac, liver and renal disease. Furthermore, we give an overview of important technical considerations necessary for clinical implementation of quantitative parametric imaging, involving data acquisition, data analysis, quality assessment, and interpretation. In order to achieve clinical implementation of these techniques, standardization of T₁ and T₂(*) mapping methodology and validation of impact on clinical decision making is needed.

Introduction

Pathological alterations in tissue composition often have similar manifestations in different organ systems such as heart, liver and kidney. To illustrate, fibrotic organs share similarities on both histopathology as imaging, including stiffness due to excessive extracellular matrix deposition, reduced vasculature, and an uneven surface due to fibroblast formation.^{1, 2} Also edema manifests in different organs as excessive fluid accumulation either within cells (cellular edema) or within the collagen matrix of the interstitial spaces (interstitial edema).³ Infiltrative diseases, e.g. iron deposition, amyloidosis, and lipid accumulation lead to systemic alterations in tissue composition causing dysfunction of different organs, including heart, liver, and kidney. These pathological changes in tissue composition can be non-invasively visualized and quantified using novel multiparametic imaging techniques, whereas conventional MR imaging only enabled qualitative image interpretation and signal intensity based analysis using arbitrary units.⁴

Direct quantification of the T₁ and T₂(*) via parametric imaging (i.e. imaging using quantitative sequences such as T₁ and T₂(*) mapping with milliseconds as the corresponding unit) addresses several of these limitations via the inherent quantitative results and elimination of user-dependent interpretation. Tissue characterization using late gadolinium enhancement (LGE) in cardiac MR is considered the gold-standard non-invasive imaging technique for the assessment of myocardial scar, however, several important limitations exist. Since LGE relies on differences in signal intensity between scar tissue and adjacent “normal” tissue, it is not sensitive for the detection of diffuse fibrosis.⁵ Additionally, signal intensities in LGE are expressed on an arbitrary scale which challenges comparison over time, and the enhancing tissues are not only influenced by technical parameters during image acquisition but also to the arbitrarily set intensity threshold.⁶ T₂ weighted imaging is commonly used to asses inflammation and edema, however, these sequences are affected by various limitations including regional differences introduced by signal variation due to phased-array coil arrays, and difficulties in differentiating edema from subendocardial blood in cardiac MR.⁷ Quantification of T₁ and T₂ values based on a quantitative pixel-wise maps can reduce the variation in assessment, and thus serve as an alternative for LGE and T₂ weighted imaging.⁸ T₁ and T₂(*) mapping not only identifies and quantifies diseased tissue contents, but also allows for direct comparison over time with reduced analysis time.⁹ Initial efforts of multiparametric imaging using T₁ and T₂(*) mapping have mainly focused on cardiac imaging, however these techniques also be applied in other organs, such as liver, and kidney. This ability of non-invasive tissue characterization could ultimately be used for better understanding of common disease pathways and monitoring the effectiveness of different therapies. An overview of potential parametric imaging methods for the assessment of different heart, liver and kidney diseases is given in Table 1. In this review, we provide an overview of potential clinical application of T₁ and T₂(*) mapping for imaging of cardiac, liver and renal disease. Furthermore, we describe important technical considerations necessary for clinical implementation of quantitative parametric imaging, involving data acquisition, data analysis, quality assessment, and interpretation.

Table 1. .

Overview of potential parametric imaging methods for the assessment of different heart, liver and kidney diseases

Parametric imaging method	Organ of interest
Heart	Liver	Kidney
Native T1	Edema (acute ischemia, acute inflammation), storage disease (amyloid, iron, lipid deposition)	Fibrosis, steatohepatitis, post-transplantation changes	Fibrosis, post-transplantation changes
ECV, Post-contrast T1	Fibrosis (replacement: chronic infarction, primary cardiomyopathy; interstitial; primary cardiomyopathy, volume overload)	Functional liver parenchyma
T2	Edema (acute ischemia, acute inflammation)	Edema (pre-clinical models only)	Edema, renal cyst progression (pre-clinical models only)
T2*	Iron deposition	Iron deposition

ECV, extracellular volume.

T₁ mapping

T₁ mapping is the geographical representation of true T₁ of certain tissues within the field of view. In order to reconstruct the T₁ map, proton spin-lattice relaxation times (T₁) are calculated for every voxel within the field of view using multiple raw images with different degrees of recovery of magnetization along the longitudinal axis following inversion recovery (IR) or saturation recovery (SR) prepulses¹⁰ (Figure 1A-B). T₁ maps are reconstructed in either colour or grey scale, where the intensity of a certain voxel represents the corresponding T₁ value. This voxel-wise T₁ mapping has led to numerous studies on the clinical utility of signal quantification for the detection of myocardial disease in cardiac MRI.¹¹ Voxel-wise T₁ mapping was first introduced by the inversion recovery based modified look-locker imaging (MOLLI) sequence,¹² and has led to the development of shortened MOLLI (shMOLLI),¹³ and variations. Other T₁ mapping acquisition techniques include SR-based sequences such as saturation-recovery single-shot acquisition,¹⁴ and mixed IR–SR combinations such as saturation-pulse prepared heart-rate independent inversion-recovery,¹⁵

Figure 1.

Magnetization inversion recovery for T₁, T₂and T₂* mapping. T1 recovery curve showing increase in the longitudinal magnetization with longer inversion times due to T₁ recovery, left curve (A). Different images are obtained following an inversion pulse at multiple different inversion times for T₁ mapping during the same phase of the cardiac cycle in subsequent heart beats (B). T₂ and T₂* recovery curves showing that as the TE increases, the myocardial signal intensity decreases due to T₂ decay, (long curve), and due to static field inhomogeneities for T₂* decay (short curve) (C). Different gradient echo images are acquired at different echo times for T₂* mapping (D), and different spin-based preparation images are acquired at different echo times for T₂ mapping (E).

T₁ mapping can be used for tissue characterization by: (a) native (non-contrast) T₁ reflecting tissue disease involving both cellular components as interstitium, or (b) extracellular volume fraction (ECV) after the administration of gadolinium based contrast agents. ECV directly quantifies the size of the extracellular space as a percentage reflecting interstitial disease, and is independent of field strength.¹⁶ ECV is calculated as follows:

$${\displaystyle ECV(\mathrm{\%})=(1-hematocrit)\times \frac{(\frac{1}{T_{1}post,tissue}-\frac{1}{T_{1}native,tissue})}{(\frac{1}{T_{1}post,bloodpool})-(\frac{1}{T_{1}native,bloodpool)}}}$$

where T₁ post is the contrast-enhanced T₁ of the tissue of interest or blood pool, T₁ tissue native is the non-enhanced T₁ of the tissue of interest or blood pool (Figure 2).

Figure 2.

Calculation of ECV. Calculation of ECV using the inverse of the signal in each pixel (1/T₁) is used to generate an R₁ map (F). The ΔR₁ map of the blood pool (ΔR₁blood) and myocardium (ΔR₁ myocard) is generated by subtracting the corresponding precontrast R₁ map from the post-contrast R₁ map. ΔR₁ map pixel values are multiplied by one minus the hematocrit level, and then divided by the mean ΔR₁blood in order to calculate ECV. The final result is a colour encoded parametric map displaying the pixel-by-pixel ECV values. ECV, extra cellular volume.

T₂ and T₂* mapping

T₂ mapping is the voxel-wise representation of the proton spin–spin relaxation time (T₂) of the tissue of interest within the field of view. T₂ values for each voxel are acquired via based T₂ weighted images at various echo times (TE) with a long repetition time in order to minimize the effect of longitudinal relaxation (Figure 1C–D). Acquired T₂ values reflect the free water content present in the tissue of interest, which can be used for quantification of edema. The most frequently used sequence for T₂ mapping is the balanced steady-state free precession (bSSFP) sequence,⁸ and other used sequences are gradient-recalled echo (GRE)¹⁷ and spiral imaging.¹⁸ These sequences are combined with several images with different T₂ preparation module TEs.

T₂ star (denoted as T₂*) mapping uses the effective T₂ value which decays faster than true T₂ due to the dephasing effects of local field inhomogeneities from susceptibility differences present within the voxel (Figure 1C–E). T₂* mapping can be used for measurement of iron content in tissues. Used T₂* mapping sequences are multiecho GRE sequences (Table 1).¹⁹

Clinical applications

Heart

Diffuse fibrosis and infiltrative cardiac diseases

One of the major advantages of T₁ mapping compared to LGE is the possibility to visualize infiltrative interstitial disease or extensive diffuse fibrosis (Figures 3 and 4). Fibrosis which is a non-physiological scarring process leading to destruction of organ architecture and organ dysfunction via excessive deposition of extracellular matrix.² Increased T₁ on native, and post-contrast images due to diffuse fibrosis has extensively been described in several diseases, such as hypertrophic cardiomyopathy, aortic stenosis, sarcoidosis, systemic sclerosis, and myocarditis²⁰ (Figure 5). Also, interstitial myocardial fibrosis after treatment with anthracycline chemotherapy has been associated with significantly increased ECV values compared with oncologic patients that had not yet initiated chemotherapy.²² These findings indicate that T₁ mapping techniques may be useful as novel risk stratification biomarkers for cardiotoxicity prior to and during treatment with anthracycline agents. Increased interstitial space does not only result from fibrosis, but may also be due to the presence of infiltrates such as in amyloidosis.^{23, 24} In amyloidosis, T₁ mapping and ECV have made great advance in diagnosing cardiac involvement and have shown to be predictive of mortality.^24–26 As such, the necessity of cardiac biopsy for confirming cardiac involvement can be debated as native T₁ and ECV can be used reliably for non-invasive diagnosis. Another exemplary disease with diffuse myocardial infiltration that can be well detected via parametric imaging is Anderson-Fabry disease. Anderson-Fabry is characterized by intracellular lysosomal lipid accumulation which results in decreased T₁ values on native T₁ mapping.^{27, 28} Other cardiomyopathies in which T₁ mapping and ECV have been described to be potentially beneficial for diagnosis are hypertrophic²⁹ and dilating cardiomyopathy,³⁰ however, further research is still needed to validate diagnostic usefulness and prognostication. Another example of an interstitial disease in which T₂* mapping can be of great value is cardiac siderosis. Previous research has showed that myocardial T₂ values correlate well with tissue iron concentration,³¹ which has enabled visualization and quantification of iron accumulation in the heart using T₂(*) mapping (Figure 6A). Parametric imaging could be besides diagnosis also be used for treatment monitoring, such as plasma cell dyscrasia suppressive agents for light-chain amyloidosis,³² enzyme replacement therapies for Anderson-Fabry,³³ and modern chelation regimes for cardiac sidersosis.³⁴ Early initiation of chelation therapy based on myocardial T₂* has drastically influenced long-term prognosis in patients with thalassemia by decreasing the annual death rate from cardiac iron overload.³⁴ When available, T₁ mapping and ECV could also be used for monitoring the effectiveness of antifibrotic treaments.³⁵

Figure 3.

Example of correspondence of ECV and LGE in a patient with PVCs with focal fibrosis. LGE shows some enhancement basal septal, which is confirmed by the ECV map constructed using the pre- and post-contrast T₁ maps. The ECV in the region of interest was 45% localized in focal septal hypertrophy, which is the likely origin of the PVC's. Quantitative T₁ and ECV maps were automatically reconstructed on a voxel-by-voxel basis after data acquisition using the T₁ map processing tool (Medis Research Edition, v. 3.0, Leiden). ECV, extra cellular volume; LGE, late gadolinium enhancement; PCV, premature ventricular contraction.

Figure 4.

Example of added value of ECV compared to LGE in a patient with familial hypertrophic cardiomyopathy with diffuse fibrosis. Non-dilated left ventricle with septal hypertrophy with diffuse fibrosis [serum hematocrit of 45%, native T₁ septum 1315 ms (N < 1350 ms), and ECV 42% (N < 35%)]. Quantitative T₁ and ECV maps were automatically reconstructed on a voxel-by-voxel basis after data acquisition using the T₁ map processing tool (Medis Research , v. 3.0, Leiden). ECV, extra cellular volume; LGE, late gadolinium enhancement.

Figure 5. .

Tissue characterization using native T₁ and ECV fraction. Absolute values for native T₁ depend greatly on field strength (1.5 or 3 T), pulse sequence (MOLLI or ShMOLLI), scanner manufacturer and post-processing. For the purpose of comparability, only studies using 1.5 T scanners were considered in this figure. Reprinted from Haaf et al²¹ publisher BioMed Central under the terms of the Creative Commons Licence. ECV, extra cellular volume; MOLLI, modified look-locker imaging; ShMOLLI, shortened MOLLI.

Figure 6.

T₂* mapping of heart (left) and liver (right) in a childhood cancer survivor at risk of secondary hemosiderosis after multiple blood transfusions and chemotherapy for acute lymphatic leukemia. Parametric imaging of heart and liver using StarQuant (Philips) heart and LiverMultiScan (Perspectum). The myocardial T₂* value was 38 ms (normal reference > 20 ms), and liver T₂* value was 13.3 ms, indicating normal T₂* values of the heart and minimal iron deposition in the liver. Quantitative T₂ maps were automatically reconstructed on a voxel-by-voxel basis after data acquisition using the T₂ map processing tool (Medis Research Edition, v. 3.0, Leiden).

Cardiac dysfunction

Functional studies have showed that higher ECV values are correlated with reduced left ventricular ejection fraction, and lower myocardial blood flow in dilated cardiomyopathy and lower systolic strain in left ventricular hypertrophy.^{29, 36} Furthermore, interstitial fibrosis in diastolic dysfunction has also been linked to the development of heart failure with preserved ejection fraction.³⁷ These findings suggest that the expansion of the extracellular matrix may be a key contributor to contractile dysfunction. Combining parametric imaging of the heart with functional cardiac MRI could be of great advantage for identifying focal areas of interstitial fibrosis that negatively influence cardiac function. There is an growing body of evidence evaluating the prognostic value of T₁ mapping and ECV in patients with cardiac dysfunction.³⁸ Several studies have been performed that evaluated the association between native T₁³⁹ and ECV^11,40–42 with incident heart failure and all-cause mortality. These studies have found that both native T₁ and ECV are more sensitive for predicting adverse events than left ventricular ejection fraction, which is the currently used for prognostication in heart failure.³⁸ However, for T₂ mapping thus far no prognostic evidence has been reported for patients with heart failure, although the diagnostic role of T₂ mapping for acute conditions such as acute myocardial infarction and acute myocarditis is promising.

Ischemic heart disease

Differentiation between acute and chronic myocardial infarction has important clinical implications. LGE, which is currently used for the detection of infarcted myocardium, is sensitive to motion artifacts, and incomplete nulling of the myocardium, and does not differentiate well between acute and chronic myocardial infarction. Early studies using T₁ mapping showed that acute and chronic myocardial infarction had different patterns of T₁ changes after the administration of gadolinium.⁴³ Besides contrast-enhanced techniques, also native T₁ and T₂ mapping have shown to be an accurate method for differentiating acute and chronic myocardial infarction via the detection of edema.^{44, 45} Expansion of current cardiac imaging protocols with T₁ and T₂ mapping could thus, potentially improve the sensitivity for the detection of myocardial infarction compared to LGE and T₂ weighted black blood imaging alone.

Myocarditis

Acute myocarditis is associated with a high mortality if untreated, however, clinical criteria alone are often of limited value for establishing the diagnosis. Both native T₁ and T₂ mapping have showed to be more sensitive for the detection of acute myocarditis with T₂ weighted and LGE MRI techniques,^{46, 47} however, native T₁ mapping was found to have a superior diagnostic performance compared with T2 mapping.⁴⁷ Moreover, recent studies have showed that both native T₁ mapping and T₂ mapping can reliably discriminate between healthy and diseased myocardial tissue,^{48, 49} and correspond to the clinical disease stage.⁵⁰ The use of LGE and ECV seems to be beneficial for the detection of more chronic stages of myocarditis.⁵⁰

Liver

Estimated annual progression rates of compensated to decompensated liver cirrhosis range between 5 and 11%,^{51, 52} and prevention of decompensation is the primary treatment goal in compensated cirrhosis.⁵³ However, currently available clinical scoring systems do not accurately identify patients at increased risk of decompensation.⁵⁴ The observation that the extent of liver enhancement by hepatobiliary specific contrast agents, such as gadobenate dimeglumine and gadoxetate disodium, is liver function dependent has led to multiple studies on contrast-enhanced T₁ mapping using these agents. Several of these studies have shown promising results indicating that hepatobiliary contrast-enhanced T₁ mapping and ECV correlates well with histological measurements of hepatic fibrosis,⁵⁵ liver function tests,^56–60 and Child-Pugh scores.⁶¹ Recent studies, however, have indicated that also native hepatic T₁ corrected for iron content (cT1) can be used for estimating liver fibrosis.^{62, 63} cT1 was found to be independently associated with survival in a proof of principle study,⁶⁴ and was not affected by the degree of adiposity or presence of ascites⁶² in contrast to other acoustic-based techniques such as elastrography.⁶² Furthermore, higher liver inflammation and fibrosis scores based on hepatic T₁ and T₂* values were found to be associated with an increased risk of liver-related adverse outcomes such as encephalopathy, ascites and liver-related death.⁶⁵

Already in 2005, it has been described that relaxation rates 1/T₂ and 1/T₂* could be used as a non-invasive method for the quantification of hepatic iron concentration, as these measures were closely correlated by iron concentration measured via liver biopsy.⁶⁶ When parametric mapping techniques became available, additional studies histologically validated the ability of T₂* mapping for the quantification hepatic iron content,^{62, 67} and assessed reproducibility.⁶² A prospective study evaluating the predictive value of T₂* on liver-related adverse outcomes found a protective effect with increasing T₂*, which is inversely related to iron load.⁶⁵ These findings are in line with previous biopsy studies that observed hepatic iron content was predictive of death in alcohol-related liver cirrhosis,⁶⁸ and more severe fibrosis in non-alcoholic fatty liver disease.⁶⁹ Non-invasive parametric imaging of the liver could ultimately contribute to personalized medicine based approaches for treatment monitoring, such as evaluating the effects of hepatic iron lowering therapy (Figure 6B)⁷⁰ or antifibrotic treatment strategies.¹ However, additional (multicenter) studies are needed in order to determine whether multiparametric MRI could indeed contribute to achieving this goal and ultimately replace liver biopsies.

Kidney

On conventional MRI of the kidney, anatomical differences between renal cortex and medulla can be clearly differentiated due to the shorter T₁ relaxation times of the cortex. Loss of this so-called corticomedullary differentiation occurs in several renal diseases and has been primarily attributed to altered T₁ relaxation times in the renal cortex.⁷¹ Recent studies suggest that characterization of renal tissue composition via true T₁ values without contrast might be useful for differentiating specific renal disease states, such as renal fibrosis imaging. Pre-clinical studies have shown that T₁ mapping could be used for the assessment of acute kidney injury and chronic kidney disease in mice.^72–74 Recent clinical studies in renal transplant patients found that renal native T₁ values correlated well with renal fibrosis severity based on histology⁷⁵ and with glomerular filtration rate after transplantation.⁷⁶ Good intra- and interexamination reproducibility has been reported for renal native T₁ mapping using the MOLLI 5 (3)3 scheme in both healthy human volunteers and diabetic nephropathy patients,⁷⁷ supporting that native T₁ could be used as a reliable and consistent measure of renal tissue composition. However, additional studies are needed to evaluate the reproducibility of renal T₁ mapping at different imaging centers with various MRI scanner manufacturers. Since native T₁ mapping is at least partially modulated by perfusion (which is also a major determinant of glomerular filtration rate), T₁ relaxation times obtained in patients with impaired renal function could theoretically be confounded by lower renal perfusion rather reflecting true fibrosis only. More research is needed to determine to what extent native renal T₁ values are affected by impaired perfusion, and whether renal native T₁ mapping has added value for clinical decision-making compared to currently available renal function markers and other MR techniques such as diffusion-weighted imaging, and blood-oxygen-level dependent imaging. Thus far, no studies have evaluated renal ECV (interstitial) using native and post-contrast T₁- mapping. The administration of contrast in patients with severely impaired renal function is controversial due to the risk of nephrogenic systemic fibrosis,⁷⁸ however, new insights suggest that modern macrocyclic GBCAs may not be associated with the development of nephrogenic systemic fibrosis even when administered to high risk chronic kidney disease patients.^79–82 Renal T₂ mapping has thus far only been evaluated in mouse models, which showed that renal cortex T₂ values increase after kidney transplantation⁷³ and that renal T₂ is highly correlated with the histological cystic index in a polycystic kidney disease model.⁸³ Further research is needed to assess whether T₂ mapping could be useful for assessment of edema, or for the prediction of cyst progression in humans.

Technical considerations for clinical implementation

Data acquisition

The decision about the used pulse sequence and parameters starts with the clinical question that needs to be answered, and the disease and organ of interest (Table 1). Roughly, it can be said that T₁ mapping can be used for imaging of fibrosis, steatosis, edema, iron without the need for contrast agents. As native T₁ is a measure of both intra- and extracellular space, it is less sensitive to increased extracellular space but more sensitive to other tissue characteristics, such as hemosiderosis, steatosis, and edema. The strength of ECV is; (a) the possibility to differentiate between intracellular vs extracellular (interstitial) compartments, and (b) its independence to field strength.⁸⁴ T₂ mapping and T₂* mapping are very sensitive for edema and hemosiderosis respectively.

Which field strength is optimal for a particular clinical application of T₁ and T₂(*) mapping is another important question. Most validation studies and references studies for cardiac parametric imaging have been performed at 1.5 T, however, most parametric imaging studies of the liver have been performed at 3 T. Advantages of higher field strengths are the increased signal-to-noise ratio, and disadvantages are the larger effects of field inhomogeneities. An overview of the advantages and disadvantages of inversion recovery versus saturation recovery based T₁ mapping techniques are presented in Table 2.

Table 2. .

Inversion recovery versus saturation recovery T₁ mapping techniques

Technique	Example	Advantages	Disadvantages
IR	MOLLI,1 shMOLLI,2 modified MOLLI	Good precision and reproducibility, few image artifacts	Less absolute accuracy
SR	SASHA3	Could potentially provide more accurate T1 measurements, less sensitive to magnetization transfer	More susceptible to noise and artifacts, reproducibility has less extensively been validated
Combined	SAPPHIRE4	Shares many of the advantages of IR and SR	Shares the disadvantages of IR

IR, inversion recovery; MOLLI, modified look-locker imaging; shMOLLI, shortened modified look-locker imaging; SASHA, saturation-recovery single-shot acquisition; SAPPIRE, saturation-pulse prepared heart-rate independent inversion-recovery; SR, saturation recovery; 

Planning

Tissues of interest should be orthogonal to the imaging plane in order to minimize through plane partial volume averaging, which is the two-chamber short axis for the heart, axial for the liver, and axial or sagittal for the kidney. Furthermore, shimming and center frequency should be adjusted to minimize off-resonance, which is especially important at higher field strengths since off-resonance variation may result in regional variations in apparent T₁.⁸⁵ Adequate breath-holding is needed for correct registration of obtained images, since misregistration can introduce substantial errors in the calculated maps. For cardiac parametric imaging, obtained images should be acquired at the same cardiac phase and respiratory position to eliminate tissue motion. Motion-correction could partly overcome the effects of suboptimal breath-holding, and minimize artifacts related to motion and misregistration. The use of fully automated motion correction and co-registration of breath-holds can significantly improve the quality of ECV maps, and increase clinical applicability.⁸⁶ New developments are the application of three-dimensional imaging and segmentation in order to achieve higher spatial resolution,⁸⁷ and the use of automated ECV measurement⁸⁶ or volumetric ECV measurement for the determination of functional liver-volume.⁸⁸

Data analysis and reporting

Clinical imaging units currently provide MR T₁ and T₂(*) mapping software that can be used for visual evaluation and basic quantification. Post-processing software with dedicated quantification packages are available, which contribute to appropriate scaling of the parametric maps in color- or greyscale to maximize differentiation between diseased and normal tissues. Regions of interest should be placed with care in order to minimize partial volume effects and should have adequate margins from tissue interfaces, such as the intracardial blood pool, pericardial fat, renal sinus fat and perirenal fat, but also large vascular and biliary structures in the liver. Quantitative error estimates in post-processing software are useful for the assessment of the reliability of measured T₁ and T₂(*) values. The availability of such quantitative error estimates are an important requirement for the use of quantitative parametric imaging in clinical decision making, since these can help to identify unreliable regions in quantitative imaging and for interpretation and for comparison of imaging protocols.⁸⁹ The importance of the quality of the pixel-wise T₁ and T₂(*) maps generated with the chosen pulse sequence, parameters, and field strength cannot be underestimated as for reliable clinical decision making high quality, artifact-free pixel-wise maps are crucial.⁸⁴ The detection of potential artifacts and handling still relies on human expertise, which hampers the easy application of these techniques in clinical practice. The Society for Cardiovascular Magnetic Resonance has recently recommended that local results in healthy volunteers for native T₁, and T₂ mapping should be primarily used and benchmarked against published reference values.⁹⁰ For clinical use, reference data based on a sufficiently large cohorts reflecting normal variations are needed. Since each T₁ and T₂(*) mapping technique has specific measurement errors, each technique should in principal be compared with normal reference values that were obtained using the same acquisition method, including same pulse sequence parameters and field strength.⁸⁴ This requires verification on whether the scanner configurations are identical to the acquisition method used in the reference studies.⁹¹ Finally, implementation of T₁ and T₂(*) mapping results into picture archiving and communication systems could facilitate and enhance the use of parametric imaging data in the clinical work environment.

Discussion

To make the transition from an investigational technique to a reliable clinical modality, T₁ and T₂(*) mapping studies need to prove that these techniques have the ability to make an early, non-invasive diagnosis or to increase confidence in a suspected diagnosis.

In order for an imaging technique to make a successful transition in clinical setting, the impact of the technique on healthcare needs to be assessed. Criteria that have been defined to assess the efficacy in diagnostic imaging are; technical feasibility, diagnostic accuracy, diagnostic impact, therapeutic impact, impact on outcome, and societal impact.⁹² Currently, cardiac T₁ mapping and hepatic T₁ and T₂* mapping fulfil the first two criteria, and an increasing amount of studies on cardiac T₁ mapping and ECV quantification have demonstrated impact on differential diagnosis, treatment strategies, and clinical outcome. Thus far, only few studies have evaluated societal impact, such as cost–benefit analysis. For multiparametric MR of the liver combined with transient elastography, it has been estimated to yield a cost saving over £500 for every patient needing diagnostic evaluation for non-alcoholic stratohepatitis.⁹³ There is an increasing need for studies evaluating to what extent T₁ and T₂(*) mapping improve diagnosis and contribute to changes in treatment strategies resulting in improved patient outcomes. In cardiac imaging, T₂* mapping is increasingly clinically used for treatment monitoring in cardiac siderosis. However considering that T₁ values overlap for the majority of cardiac pathologies the value of T₁ mapping beyond conventional sequences for diagnostic purposes remains to be proven. Since hepatic steatosis and siderosis can be easily and accurately quantified by parametric imaging and enable treatment response evaluation, it is expected that T₁ and T₂* mapping will be increasingly used clinically in the near future. Parametric imaging of the kidney, however, has just recently entered the research phase. An overview of potential clinical applications of T₁, ECV, T₂, and T₂* is given in Figure 7 . Additional to the above mentioned criteria, more studies are needed to provide good reference data for T₁ and T₂(*) mapping in order to introduce these techniques into clinical practice.

Figure 7.

Typical appearance of T₁, ECV, T₂, and T₂* maps in heart, liver, and kidney of healthy subjects (upper row) and in patients with myocardial and liver disease (second to fourth row) (Medis Research Edition, v. 3.0, Leiden). Adapted by permission from BioMed Central under the terms of the Creative Commons Licence,⁹⁰ and adapted by permission from BMJ Publishing Group Limited.⁵⁵ECV, extra cellular volume.

Ultimately, the intra- and interexamination reproducibility of measured T₁ and T₂(*) values determines the clinical utility of pixel-wise T₁ and T₂(*) mapping for disease assessment. To be of clinical value, assessed experimental and biologic variation in the quantified T₁ and T₂(*) values should be smaller than the changes caused by disease. In order to assess this, sufficiently large cohorts of subjects are needed to guaranty the robustness of a classifier (e.g. sensitivity and specificity) and ultimately, findings should be validated in a multicentre trial. Two large ongoing multicentre studies on this topic are currently registered on ClinicalTrials.gov. One will evaluate whether myocardial fibrosis based on LGE and T₁ mapping can predict all cause and cardiovascular mortality, with an aimed sample size of 1500 participants.⁹⁴ The second study aims investigates whether it is cost-effective to use T₁ and T₂* imaging of the liver as a standardized diagnostic test for liver disease in 2000 participants.⁶⁴ The outcomes of these studies contribute to determining whether parametric imaging will truly find its way into clinical practice, or whether it will remain considered as an “investigational technique” by medical professionals, and healthcare institutions.

In conclusion, T₁ and T₂(*) mapping can be considered promising techniques that can be used in addition to conventional MRI for the quantification of pathological changes in tissue composition. Disease entities for which T₁ and T₂(*) mapping could be used clinically are cardiomyopathies, and ischemic heart disease, and other possible applications are the quantification of liver cirrhosis, hemosiderosis and renal fibrosis. Availability of normative data together with standardization of data acquisition, and analysis is warranted. Multicenter trials with sufficient sample size are needed to establish the impact of T₁ and T₂(*) mapping on clinical outcome and economic benefit.