Multiecho water-fat separation and simultaneous R2* estimation with multifrequency fat spectrum modeling

Diurnal Variation of Proton Density Fat Fraction in the Liver Using Quantitative Chemical Shift Encoded MRI

BACKGROUND

Whole-organ, noninvasive techniques for the detection and quantification of nonalcoholic fatty liver disease features have clinical and research applications. However, the effect of time of day, hydration status, and meals are unknown factors with potential to impact bias, precision, reproducibility, and repeatability of chemical shift-encoded MRI (CSE-MRI) to quantify liver proton density fat fraction (PDFF).

PURPOSE

To assess the effect of diurnal variation on PDFF using CSE-MRI, including the effect of time of day, the effect of meals and hydration status, as well as the day to day variability.

STUDY TYPE

Prospective.

SUBJECTS

Eleven healthy subjects and nine patients with observed hepatic steatosis.

FIELD STRENGTH/SEQUENCES

A commercial quantitative confounder-corrected CSE-MRI sequence (IDEAL IQ) and an MR spectroscopy (MRS) sequence (multiecho STEAM) were acquired at 1.5T.

ASSESSMENT

MRI-PDFF and MRS-PDFF estimates were compared across six visits (before and after a controlled breakfast, before and after an uncontrolled lunch, at approximately 4 pm, and then before breakfast on the following day) with three repeated measures for a total of 360 MRI-PDFF and MRS-PDFF measurements.

STATISTICAL TESTS

Linear regression, Bland-Altman analysis, and mixed effect models were used to determine the bias, precision, and repeatability of PDFF measurements.

RESULTS

No statistically significant linear trend was observed across visits for either MRI-PDFF or MRS-PDFF (P = 0.31 and 0.37, respectively). The repeatability was measured to be 0.86% for MRI-PDFF and 1.1% for MRS-PDFF over all six visits. For MRI-PDFF, the variability between all six visits (0.94%) was only slightly higher than within each visit (0.66%), with P < 0.001. For MRS-PDFF, the variability between all six visits was 1.29%, compared with 0.87% within each visit (P < 0.001).

DATA CONCLUSION

Our results may indicate that it is not necessary to control for the time of day or the fasting/fed state of the patient when measuring PDFF using CSE-MRI.

LEVEL OF EVIDENCE

2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2020;51:407-414.

Evaluation of hepatic iron concentration heterogeneities using the MRI R2* mapping method

OBJECTIVE

To measure hepatic iron concentration (HIC) heterogeneities using a magnetic resonance R2* mapping method.

PATIENTS AND METHODS

Ninety-four patients with suspected hepatic iron overload and 10 volunteers were included prospectively. A multi-echo R2* sequence with fat saturation and with three post-processing fitting methods (a single exponential decay model with or without truncation, SED and SEDt, and a constant offset model, COS) was compared to a signal intensity ratio method (SIR), considered as the reference. HIC heterogeneity was evaluated from R2* mapping after placing a ROI on each liver segment.

RESULTS

A strong linear correlation between SIR and R2* methods using the SEDt and COS models was observed (r = 0.973 and 0.955, respectively). Volunteers and patient liver variabilities, quantified by mean intra-liver standard deviation (SD) were 1.58 μmol/g (mean range 5.06 μmol/g) and 4.73 μmol/g (mean range 19.08 μmol/g), respectively. For the patient group, the highest HIC was observed in the IV^th segment. Heterogeneity increased for patients with an HIC > 60 μmol/g (mean intra-liver SD = 13.90 μmol/g; mean range = 50.60 μmol/g).

CONCLUSION

This study is the first to demonstrate in vivo HIC heterogeneities using whole-liver mapping analysis. These preliminary results require confirmation through further studies, but might be useful in cases of single ROI analysis.

Magnetic Resonance Imaging: A Personalized Approach to Understanding Fatty Liver Disease

OBJECTIVES

To assess information retention by parents/caretakers regarding nonalcoholic fatty liver disease (NAFLD) utilizing the actual image of their child's affected liver.

METHODS

In this pilot study, parents/caretakers of children with newly diagnosed NAFLD were presented with an magnetic resonance (MR) image of their child's fatty liver. An adjacent image of a normal-appearing liver was used to highlight the degree of fat accumulation present in their child's liver. The appearance of the fatty liver was used as an adjunct to patient education as provided by a nurse clinician. The efficacy of this approach was determined by a set of image- and disease-specific queries. Health literacy was assessed concurrently by the Newest Vital Sign (NVS) instrument. The image- and disease-specific queries were then repeated by telephone follow-up 2 to 4 weeks after initial clinic visit.

RESULTS

Parents/caretakers initially gave 100% correct responses regarding the variation of appearance of normal liver (pink) and their child's fatty liver (yellow). They also all correctly stated the fat content initially. At follow-up, their recall was 95% for the appearance of normal liver and 81% for fatty liver; recall was only 52% for fat content at follow-up. Nonvisualized elements of nonalcoholic steatohepatitis (NASH) and cirrhosis were not identified or recalled as well. Results may have been influenced by parent/caretaker health literacy competence.

CONCLUSIONS

Personalized images of fatty liver were effective visualization tools for parents/caretakers to comprehend NAFLD and comprehension was not compromised by health literacy. Clear visual instruments may improve parent/caretaker comprehension of these conditions and may help to address deficiencies in health literacy.

Separation of water and fat signal in whole-body gradient echo scans using convolutional neural networks

Purpose

To perform and evaluate water–fat signal separation of whole‐body gradient echo scans using convolutional neural networks.

Methods

Whole‐body gradient echo scans of 240 subjects, each consisting of 5 bipolar echoes, were used. Reference fat fraction maps were created using a conventional method. Convolutional neural networks, more specifically 2D U‐nets, were trained using 5‐fold cross‐validation with 1 or several echoes as input, using the squared difference between the output and the reference fat fraction maps as the loss function. The outputs of the networks were assessed by the loss function, measured liver fat fractions, and visually. Training was performed using a graphics processing unit (GPU). Inference was performed using the GPU as well as a central processing unit (CPU).

Results

The loss curves indicated convergence, and the final loss of the validation data decreased when using more echoes as input. The liver fat fractions could be estimated using only 1 echo, but results were improved by use of more echoes. Visual assessment found the quality of the outputs of the networks to be similar to the reference even when using only 1 echo, with slight improvements when using more echoes. Training a network took at most 28.6 h. Inference time of a whole‐body scan took at most 3.7 s using the GPU and 5.8 min using the CPU.

Conclusion

It is possible to perform water–fat signal separation of whole‐body gradient echo scans using convolutional neural networks. Separation was possible using only 1 echo, although using more echoes improved the results.

Dynamic water/fat separation and B 0 inhomogeneity mapping-joint estimation using undersampled triple-echo multi-spoke radial FLASH

PURPOSE

To achieve dynamic water/fat separation and B 0 field inhomogeneity mapping via model-based reconstructions of undersampled triple-echo multi-spoke radial FLASH acquisitions.

METHODS

This work introduces an undersampled triple-echo multi-spoke radial FLASH sequence, which uses (i) complementary radial spokes per echo train for faster spatial encoding, (ii) asymmetric echoes for flexible and nonuniform echo spacing, and (iii) a golden angle increment across frames for optimal k-space coverage. Joint estimation of water, fat, B 0 inhomogeneity, and coil sensitivity maps from undersampled triple-echo data poses a nonlinear and non-convex inverse problem which is solved by a model-based reconstruction with suitable regularization. The developed methods are validated using phantom experiments with different degrees of undersampling. Real-time MRI studies of the knee, liver, and heart are conducted without prospective gating or retrospective data sorting at temporal resolutions of 70, 158, and 40 ms, respectively.

RESULTS

Up to 18-fold undersampling is achieved in this work. Even in the presence of rapid physiological motion, large B 0 field inhomogeneities, and phase wrapping, the model-based reconstruction yields reliably separated water/fat maps in conjunction with spatially smooth inhomogeneity maps.

CONCLUSIONS

The combination of a triple-echo acquisition and joint reconstruction technique provides a practical solution to time-resolved and motion robust water/fat separation at high spatial and temporal resolution.

Proton density fat fraction MRI of vertebral bone marrow: Accuracy, repeatability, and reproducibility among readers, field strengths, and imaging platforms

BACKGROUND

Chemical shift-encoding based water-fat MRI is an emerging method to noninvasively assess proton density fat fraction (PDFF), a promising quantitative imaging biomarker for estimating tissue fat concentration. However, in vivo validation of PDFF is still lacking for bone marrow applications.

PURPOSE

To determine the accuracy and precision of MRI-determined vertebral bone marrow PDFF among different readers and across different field strengths and imager manufacturers.

STUDY TYPE

Repeatability/reproducibility.

SUBJECTS

Twenty-four adult volunteers underwent lumbar spine MRI with one 1.5T and two different 3.0T MR scanners from two vendors on the same day.

FIELD STRENGTH/SEQUENCE

1.5T and 3.0T/3D spoiled-gradient echo multipoint Dixon sequences.

ASSESSMENT

Two independent readers measured intravertebral PDFF for the three most central slices of the L1-5 vertebral bodies. Single-voxel MR spectroscopy (MRS)-determined PDFF served as the reference standard for PDFF estimation.

STATISTICAL TESTS

Accuracy and bias were assessed by Pearson correlation, linear regression analysis, and Bland-Altman plots. Repeatability and reproducibility were evaluated by Wilcoxon signed rank test, Friedman test, and coefficients of variation. Intraclass correlation coefficients were used to validate intra- and interreader as well as intraimager agreements.

RESULTS

MRI-based PDFF estimates of lumbar bone marrow were highly correlated (r² = 0.899) and accurate (mean bias, -0.6%) against the MRS-determined PDFF reference standard. PDFF showed high linearity (r² = 0.972-0.978) and small mean bias (0.6-1.5%) with 95% limits of agreement within ±3.4% across field strengths, imaging platforms, and readers. Repeatability and reproducibility of PDFF were high, with the mean overall coefficient of variation being 0.86% and 2.77%, respectively. The overall intraclass correlation coefficient was 0.986 as a measure for an excellent interreader agreement.

DATA CONCLUSION

MRI-based quantification of vertebral bone marrow PDFF is highly accurate, repeatable, and reproducible among readers, field strengths, and MRI platforms, indicating its robustness as a quantitative imaging biomarker for multicentric studies.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:1762-1772.

Prevalence of Nonalcoholic Fatty Liver Disease in Children with Obesity

OBJECTIVES

To determine the prevalence of nonalcoholic fatty liver disease (NAFLD) in children with obesity because current estimates range from 1.7% to 85%. A second objective was to evaluate the diagnostic accuracy of alanine aminotransferase (ALT) for NAFLD in children with obesity.

STUDY DESIGN

We evaluated children aged 9-17 years with obesity for the presence of NAFLD. Diseases other than NAFLD were excluded by history and laboratories. Hepatic steatosis was measured by liver magnetic resonance imaging proton density fat fraction. The diagnostic accuracy of ALT for detecting NAFLD was evaluated.

RESULTS

The study included 408 children with obesity that had a mean age of 13.2 years and mean body mass index percentile of 98.0. The study population had a mean ALT of 32 U/L and median hepatic magnetic resonance imaging proton density fat fraction of 3.7%. The estimated prevalence of NAFLD was 26.0% (95% CI 24.2%-27.7%), 29.4% in male patients (CI 26.1%-32.7%) and 22.6% in female patients (CI 16.0%-29.1%). Optimal ALT cut-point was 42 U/L (47.8% sensitivity, 93.2% specificity) for male and 30 U/L (52.1% sensitivity, 88.8% specificity) for female patients. The classification and regression tree model with sex, ALT, and insulin had 80% diagnostic accuracy for NAFLD.

CONCLUSIONS

NAFLD is common in children with obesity, but NAFLD and obesity are not concomitant. In children with obesity, NAFLD is present in nearly one-third of boys and one-fourth of girls.

Noninvasive evaluation of nonalcoholic fatty liver disease: Current evidence and practice

With the increasing number of individuals with diabetes and obesity, nonalcoholic fatty liver disease (NAFLD) is becoming increasingly prevalent, affecting one-quarter of adults worldwide. The spectrum of NAFLD ranges from simple steatosis or nonalcoholic fatty liver (NAFL) to nonalcoholic steatohepatitis (NASH). NAFLD, especially NASH, may progress to fibrosis, leading to cirrhosis and hepatocellular carcinoma. NAFLD can impose a severe economic burden, and patients with NAFLD-related terminal or deteriorative liver diseases have become one of the main groups receiving liver transplantation. The increasing prevalence of NAFLD and the severe outcomes of NASH make it necessary to use effective methods to identify NAFLD. Although recognized as the gold standard, biopsy is limited by its sampling bias, poor acceptability, and severe complications, such as mortality, bleeding, and pain. Therefore, noninvasive methods are urgently needed to avoid biopsy for diagnosing NAFLD. This review discusses the current noninvasive methods for assessing NAFLD, including steatosis, NASH, and NAFLD-related fibrosis, and explores the advantages and disadvantages of measurement tools. In addition, we analyze potential noninvasive biomarkers for tracking disease processes and monitoring treatment effects, and explore effective algorithms consisting of imaging and nonimaging biomarkers for diagnosing advanced fibrosis and reducing unnecessary biopsies in clinical practice.

Magnitude-intrinsic water-fat ambiguity can be resolved with multipeak fat modeling and a multipoint search method

Purpose

To develop a postprocessing algorithm for multiecho chemical‐shift encoded water–fat separation that estimates proton density fat fraction (PDFF) maps over the full dynamic range (0‐100%) using multipeak fat modeling and multipoint search optimization. To assess its accuracy, reproducibility, and agreement with state‐of‐the‐art complex‐based methods, and to evaluate its robustness to artefacts in abdominal PDFF maps.

Methods

We introduce MAGO (MAGnitude‐Only), a magnitude‐based reconstruction that embodies multipeak liver fat spectral modeling and multipoint optimization, and which is compatible with asymmetric echo acquisitions. MAGO is assessed first for accuracy and reproducibility on publicly available phantom data. Then, MAGO is applied to N = 178 UK Biobank cases, in which its liver PDFF measures are compared using Bland‐Altman analysis with those from a version of the hybrid iterative decomposition of water and fat with echo asymmetry and least squares estimation (IDEAL) algorithm, LiverMultiScan IDEAL (LMS IDEAL, Perspectum Diagnostics Ltd, Oxford, UK). Finally, MAGO is tested on a succession of high field challenging cases for which LMS IDEAL generated artefacts in the PDFF maps.

Results

Phantom data showed accurate, reproducible MAGO PDFF values across manufacturers, field strengths, and acquisition protocols. Moreover, we report excellent agreement between MAGO and LMS IDEAL for 6‐echo, 1.5 tesla human acquisitions (bias = −0.02% PDFF, 95% confidence interval = ±0.13% PDFF). When tested on 12‐echo, 3 tesla cases from different manufacturers, MAGO was shown to be more robust to artefacts compared to LMS IDEAL.

Conclusion

MAGO resolves the water–fat ambiguity over the entire fat fraction dynamic range without compromising accuracy, therefore enabling robust PDFF estimation where phase data is inaccessible or unreliable and complex‐based and hybrid methods fail.

Fatty infiltration in the thigh muscles in knee osteoarthritis: a systematic review and meta-analysis

Knee osteoarthritis is a chronic degenerative joint disease, influenced by inflammatory, mechanical and metabolic processes. Current literature shows that thigh muscles of people with knee osteoarthritis can have increased infiltration of fat, both between and within the muscles (inter- and intramuscular fat). The fatty infiltration in the thigh in this population is correlated to systemic inflammation, poor physical function, and muscle impairment and leads to metabolic impairments and muscle disfunction. The objective of this study is to systematically review the literature comparing the amount of fatty infiltration between people with knee osteoarthritis and healthy controls. A literature search on the databases MEDLINE, Embase, CINAHL SPORTDiscuss, Web of Science and Scopus from insertion to December 2018, resulted in 1035 articles, from which 7 met inclusion/exclusion criteria and were included in the review. All included studies analyzed the difference in intermuscular fat and only one study analyzed intramuscular fat. A meta-analysis (random effects model) transforming data into standardized mean difference was performed for intermuscular fat (six studies). The meta-analysis showed a standardized mean difference of 0.39 (95% confidence interval from 0.25 to 0.53), showing that people with knee osteoarthritis have more intermuscular fat than healthy controls. The single study analyzing intramuscular fat shows that people with knee osteoarthritis have more intramuscular fat fraction than healthy controls. People with knee osteoarthritis have more fatty infiltration around the thigh than people with no knee osteoarthritis. That conclusion is stronger for intermuscular fat than intramuscular fat, based on the quality and number of studies analyzed.

Water/fat separation for distortion-free EPI with point spread function encoding

PURPOSE

Effective removal of chemical-shift artifacts in echo-planar imaging (EPI) is a challenging problem especially with severe field inhomogeneity. This study aims to develop a reliable water/fat separation technique for point spread function (PSF) encoded EPI (PSF-EPI) by using its intrinsic multiple echo-shifted images.

THEORY AND METHODS

EPI with PSF encoding can achieve distortion-free imaging and can be highly accelerated using the tilted-CAIPI technique. In this study, the chemical-shift encoding existing in the intermediate images with different time shifts of PSF-EPI is used for water/fat separation, which is conducted with latest water/fat separation algorithms. The method was tested in T1-weighted, T2-weighted, and diffusion weighted imaging in healthy volunteers.

RESULTS

The ability of the proposed method to separate water/fat using intrinsic PSF-EPI signals without extra scans was demonstrated through in vivo T1-weighted, T2-weighted, and diffusion weighted imaging experiments. By exploring different imaging contrasts and regions, the results show that this PSF-EPI based method can separate water/fat and remove fat residues robustly.

CONCLUSION

By using the intrinsic signals of PSF-EPI for water/fat separation, fat signals can be effectively suppressed in EPI even with severe field inhomogeneity. This water/fat separation method for EPI can be extended to multiple image contrasts. The distortion-free PSF-EPI technique, thus, has the potential to provide anatomical and functional images with high-fidelity and practical acquisition efficiency.

Resolving estimation uncertainties of chemical shift encoded fat-water imaging using magnetization transfer effect

PURPOSE

B₀ field inhomogeneity may cause significant errors in chemical shift encoding-based fat-water (F/W) separation. We describe a new approach to improve its robustness using novel B₀ field map pre-estimation.

METHODS

Our method exploits insensitivity of fat to magnetization transfer effect, which allows generating fat-insensitive B₀ field priors with full or partial spatial support using a low-resolution magnetization transfer-weighted scan. The full prior can be employed by most F/W separation methods for initialization or data demodulation. We also propose a modified region-growing algorithm in which the partial prior is utilized for its initial seeding.

RESULTS

The magnetization transfer-based B₀ priors significantly reduced F/W errors of three representative F/W separation methods in all cases. In cases with moderate B₀ inhomogeneity, the full prior allowed error-free separation even with basic, voxel-independent processing. When coupled with methods exploiting B₀ field smoothness, it significantly improved separation accuracy even in the presence of strong inhomogeneities. Seeding the region-growing with the partial prior significantly improved performance of F/W separation, including cases with spatially disconnected tissues.

CONCLUSION

Magnetization transfer-based B₀ field pre-estimation provides valuable prior information for F/W separation, which may significantly improve its robustness at the expense of nominal (< 5%-10%) scan time increase.

Marrow adiposity as an indicator for insulin resistance in postmenopausal women with newly diagnosed type 2 diabetes - an investigation by chemical shift-encoded water-fat MRI

BACKGROUND

Marrow fat accumulates in diabetic conditions but remains elusive. The published works on the relationships between marrow fat phenotypes and glucose homeostasis are controversial.

PURPOSE

To detect the association of insulin resistance with marrow adiposity in postmenopausal women with newly diagnosed type 2 diabetes (T2D) using chemical shift-encoded water-fat MRI.

METHODS

We measured vertebral proton density fat fraction (PDFF) by 3T-MRI in 75 newly diagnosed T2D and 20 nondiabetic postmenopausal women. Bone mineral density (BMD), whole body fat mass and lean mass were determined by dual-energy X-ray absorptiometry. Insulin sensitivity was estimated using the homeostasis model assessment of insulin resistance (HOMA-IR).

RESULTS

Lumbar spine PDFF was higher in women with T2D (65.9 ± 6.8%) than those without diabetes (59.5 ± 6.1%, P = 0.009). There was a consistent inverse association between the vertebral PDFF and BMD. PDFF had a positive association with glycated hemoglobin and HOMA-IR but not with fasting plasma glucose and insulin. PDFF was significantly increased, and BMD was decreased in a linear trend from the lowest (<1.90) to highest (≥2.77) HOMA-IR quartile. Multivariate linear regression analyses revealed a positive association between log-transformed HOMA-IR and PDFF after adjustment for multiple covariates (ß = 0.382, P < 0.001). The positive association of HOMA-IR with PDFF remained robust when total body lean mass and fat mass, BMD was entered into the multivariate regression model, respectively (ß = 0.293 and ß = 0.251, respectively; all P <0.05).

CONCLUSIONS

Elevated HOMA-IR was linked to higher marrow fat fraction in postmenopausal women with newly diagnosed T2D independently of body compositions.

Mapping tissue water T1 in the liver using the MOLLI T1 method in the presence of fat, iron and B0 inhomogeneity

Modified Look-Locker inversion recovery (MOLLI) T1 mapping sequences can be useful in cardiac and liver tissue characterization, but determining underlying water T1 is confounded by iron, fat and frequency offsets. This article proposes an algorithm that provides an independent water MOLLI T1 (referred to as on-resonance water T1 ) that would have been measured if a subject had no fat and normal iron, and imaging had been done on resonance. Fifteen NiCl2 -doped agar phantoms with different peanut oil concentrations and 30 adults with various liver diseases, nineteen (63.3%) with liver steatosis, were scanned at 3 T using the shortened MOLLI (shMOLLI) T1 mapping, multiple-echo spoiled gradient-recalled echo and 1 H MR spectroscopy sequences. An algorithm based on Bloch equations was built in MATLAB, and water shMOLLI T1 values of both phantoms and human participants were determined. The quality of the algorithm's result was assessed by Pearson's correlation coefficient between shMOLLI T1 values and spectroscopically determined T1 values of the water, and by linear regression analysis. Correlation between shMOLLI and spectroscopy-based T1 values increased, from r = 0.910 (P < 0.001) to r = 0.998 (P < 0.001) in phantoms and from r = 0.493 (for iron-only correction; P = 0.005) to r = 0.771 (for iron, fat and off-resonance correction; P < 0.001) in patients. Linear regression analysis revealed that the determined water shMOLLI T1 values in patients were independent of fat and iron. It can be concluded that determination of on-resonance water (sh)MOLLI T1 independent of fat, iron and macroscopic field inhomogeneities was possible in phantoms and human subjects.

Renal and adrenal masses containing fat at MRI: Proposed nomenclature by the society of abdominal radiology disease-focused panel on renal cell carcinoma

This article proposes a consensus nomenclature for fat-containing renal and adrenal masses at MRI to reduce variability, improve understanding, and enhance communication when describing imaging findings. The MRI appearance of "macroscopic fat" occurs due to a sufficient number of aggregated adipocytes and results in one or more of: 1) intratumoral signal intensity (SI) loss using fat-suppression techniques, or 2) chemical shift artifact of the second kind causing linear or curvilinear India-ink (etching) artifact within or at the periphery of a mass at macroscopic fat-water interfaces. "Macroscopic fat" is most commonly observed in adrenal myelolipoma and renal angiomyolipoma (AML) and only rarely encountered in other adrenal cortical tumors and renal cell carcinomas (RCC). Nonlinear noncurvilinear signal intensity loss on opposed-phase (OP) compared with in-phase (IP) chemical shift MRI (CSI) may be referred to as "microscopic fat" and is due to: a) an insufficient amount of adipocytes, or b) the presence of fat within tumor cells. Determining whether the signal intensity loss observed on CSI is due to insufficient adipocytes or fat within tumor cells cannot be accomplished using CSI alone; however, it can be inferred when other imaging features strongly suggest a particular diagnosis. Fat-poor AML are homogeneously hypointense on T₂ -weighted (T₂ W) imaging and avidly enhancing; signal intensity loss at OP CSI is uncommon, but when present is usually focal and is caused by an insufficient number of adipocytes within adjacent voxels. Conversely, clear-cell RCC are heterogeneously hyperintense on T₂ W imaging and avidly enhancing, with the signal intensity loss observed on OP CSI being typically diffuse and due to fat within tumor cells. Adrenal adenomas, adrenal cortical carcinoma, and adrenal metastases from fat-containing primary malignancies also show signal intensity loss on OP CSI due to fat within tumor cells and not from intratumoral adipocytes. Level of Evidence: 5 Technical Efficacy Stage: 3 J. Magn. Reson. Imaging 2019;49:917-926.

Differentiating supraclavicular from gluteal adipose tissue based on simultaneous PDFF and T2 * mapping using a 20-echo gradient-echo acquisition

Background

Adipose tissue (AT) can be classified into white and brown/beige subtypes. Chemical shift encoding‐based water–fat MRI‐techniques allowing simultaneous mapping of proton density fat fraction (PDFF) and T₂* result in a lower PDFF and a shorter T₂* in brown compared with white AT. However, AT T₂* values vary widely in the literature and are primarily based on 6‐echo data. Increasing the number of echoes in a multiecho gradient‐echo acquisition is expected to increase the precision of AT T₂* mapping.

Purpose

1) To mitigate issues of current T₂*‐measurement techniques through experimental design, and 2) to investigate gluteal and supraclavicular AT T₂* and PDFF and their relationship using a 20‐echo gradient‐echo acquisition.

Study Type

Prospective.

Subjects

Twenty‐one healthy subjects.

Field Strength/Sequence Assessment

First, a ground truth signal evolution was simulated from a single‐T₂* water–fat model. Second, a time‐interleaved 20‐echo gradient‐echo sequence with monopolar gradients of neck and abdomen/pelvis at 3 T was performed in vivo to determine supraclavicular and gluteal PDFF and T₂*. Complex‐based water–fat separation was performed for the first 6 echoes and the full 20 echoes. AT depots were segmented.

Statistical Tests

Mann‐Whitney test, Wilcoxon signed‐rank test and simple linear regression analysis.

Results

Both PDFF and T₂* differed significantly between supraclavicular and gluteal AT with 6 and 20 echoes (PDFF: P < 0.0001 each, T₂*: P = 0.03 / P < 0.0001 for 6/20 echoes). 6‐echo T₂* demonstrated higher standard deviations and broader ranges than 20‐echo T₂*. Regression analyses revealed a strong relationship between PDFF and T₂* values per AT compartment (R² = 0.63 supraclavicular, R² = 0.86 gluteal, P < 0.0001 each).

Data Conclusion

The present findings suggest that an increase in the number of sampled echoes beyond 6 does not affect AT PDFF quantification, whereas AT T₂* is considerably affected. Thus, a 20‐echo gradient‐echo acquisition enables a multiparametric analysis of both AT PDFF and T₂* and may therefore improve MR‐based differentiation between white and brown fat.

Level of Evidence: 2

Technical Efficacy: Stage 2

J. Magn. Reson. Imaging 2019;50:424–434.

Reliability of measuring the fat content of the lumbar vertebral marrow and paraspinal muscles using MRI mDIXON-Quant sequence

PURPOSE

We aimed to assess the reliability of measuring the fat content of the lumbar vertebral marrow and the paraspinal muscles using magnetic resonance imaging (MRI) mDIXON-Quant sequence.

METHODS

Thirty-one healthy volunteers were included. All participants underwent liver mDIXON-Quant imaging on a 3.0 T Philips MRI scanner by observer A. Within two weeks, observer B repeated the scan. After the examination, each observer independently measured the fat content of the third lumbar vertebra (L3), and the psoas (PS), erector spinae (ES), and multifidus (MF) muscles on central L3 axial images. After two weeks, each observer repeated the same measurements. They were blinded to their previous results. Reliability was estimated by evaluating the repeatability and reproducibility.

RESULTS

The repeatability of the fat content measurements of L3, PS, ES, and MF was high. The intraclass correlation coefficients of the fat content of L3, PS, ES, and MF were 0.997, 0.984, 0.997, and 0.995 for observer A and 0.948, 0.974, 0.963, and 0.995 for observer B, respectively. The reproducibility of the measurement of the fat content of L3, PS, ES, and MF was high, and the interclass correlation coefficients were 0.984, 0.981, 0.977, and 0.998, respectively.

CONCLUSION

Using mDIXON-Quant imaging to measure the fat content of the lumbar vertebral marrow and paraspinal muscles shows high reliability and is suitable for use in clinical practice.

Marrow adipose tissue imaging in humans

Bone strength is affected not only by bone mineral density (BMD) and bone microarchitecture but also its microenvironment. Recent studies have focused on the role of marrow adipose tissue (MAT) in the pathogenesis of bone loss. Osteoblasts and adipocytes arise from a common mesenchymal stem cell within bone marrow and many osteoporotic states, including aging, medication use, immobility, over - and undernutrition are associated with increased marrow adiposity. Advancements in imaging technology allow the non-invasive quantification of MAT. This article will review magnetic resonance imaging (MRI)- and computed tomography (CT)-based imaging technologies to assess the amount and composition of MAT. The techniques that will be discussed are anatomic T1-weighted MRI, water-fat imaging, proton MR spectroscopy, single energy CT and dual energy CT. Clinical applications of MRI and CT techniques to determine the role of MAT in patients with obesity, anorexia nervosa, and type 2 diabetes will be reviewed.

Fat Quantification in the Vertebral Body: Comparison of Modified Dixon Technique with Single-Voxel Magnetic Resonance Spectroscopy

Objective

To compare the lumbar vertebral bone marrow fat-signal fractions obtained from six-echo modified Dixon sequence (6-echo m-Dixon) with those from single-voxel magnetic resonance spectroscopy (MRS) in patients with low back pain.

Materials and Methods

Vertebral bone marrow fat-signal fractions were quantified by 6-echo m-Dixon (repetition time [TR] = 7.2 ms, echo time (TE) = 1.21 ms, echo spacing = 1.1 ms, total imaging time = 50 seconds) and single-voxel MRS measurements in 25 targets (23 normal bone marrows, two focal lesions) from 24 patients. The point-resolved spectroscopy sequence was used for localized single-voxel MRS (TR = 3000 ms, TE = 35 ms, total scan time = 1 minute 42 seconds). A 2 × 2 × 1.5 cm³ voxel was placed within the normal L2 or L3 vertebral body, or other lesions including a compression fracture or metastasis. The bone marrow fat spectrum was characterized on the basis of the magnitude of measurable fat peaks and a priori knowledge of the chemical structure of triglycerides. The imaging-based fat-signal fraction results were then compared to the MRS-based results.

Results

There was a strong correlation between m-Dixon and MRS-based fat-signal fractions (slope = 0.86, R² = 0.88, p < 0.001). In Bland-Altman analysis, 92.0% (23/25) of the data points were within the limits of agreement. Bland-Altman plots revealed a slight but systematic error in the m-Dixon based fat-signal fraction, which showed a prevailing overestimation of small fat-signal fractions (< 20%) and underestimation of high fat-signal fractions (> 20%).

Conclusion

Given its excellent agreement with single-voxel-MRS, 6-echo m-Dixon can be used for visual and quantitative evaluation of vertebral bone marrow fat in daily practice.

Monitoring Fatty Liver Disease with MRI Following Bariatric Surgery: A Prospective, Dual-Center Study

Purpose To longitudinally monitor liver fat before and after bariatric surgery by using quantitative chemical shift-encoded (CSE) MRI and to compare with changes in body mass index (BMI), weight, and waist circumference (WC). Materials and Methods For this prospective study, which was approved by the internal review board, a total of 126 participants with obesity who were undergoing evaluation for bariatric surgery with preoperative very low calorie diet (VLCD) were recruited from June 27, 2010, through May 5, 2015. Written informed consent was obtained from all participants. Participants underwent CSE MRI measuring liver proton density fat fraction (PDFF) before VLCD (2-3 weeks before surgery), after VLCD (1-3 days before surgery), and 1, 3, and 6-10 months following surgery. Linear regression was used to estimate rates of change of PDFF (ΔPDFF) and body anthropometrics. Initial PDFF (PDFF₀), initial anthropometrics, and anthropometric rates of change were evaluated as predictors of ΔPDFF. Mixed-effects regression was used to estimate time to normalization of PDFF. Results Fifty participants (mean age, 51.0 years; age range, 27-70 years), including 43 women (mean age, 50.8 years; age range, 27-70 years) and seven men (mean age, 51.7 years; age range, 36-62 years), with mean PDFF₀ ± standard deviation of 18.1% ± 8.6 and mean BMI₀ of 44.9 kg/m² ± 6.5 completed the study. By 6-10 months following surgery, mean PDFF decreased to 4.9% ± 3.4 and mean BMI decreased to 34.5 kg/m² ± 5.4. Mean estimated time to PDFF normalization was 22.5 weeks ± 11.5. PDFF₀ was the only strong predictor for both ΔPDFF and time to PDFF normalization. No body anthropometric correlated with either outcome. Conclusion Average liver proton density fat fraction (PDFF) decreased to normal (< 5%) by 6-10 months following surgery, with mean time to normalization of approximately 5 months. Initial PDFF was a strong predictor of both rate of change of PDFF and time to normalization. Body anthropometrics did not predict either outcome. Online supplemental material is available for this article. © RSNA, 2018.

Measurement and correction of the bulk magnetic susceptibility effects of fat: application in venous oxygen saturation imaging

PURPOSE

To develop a correction method for the effects of the magnetic susceptibility of fat (χ_Fat ) on the calculation of venous oxygen saturation (SvO₂ ).

THEORY

The magnetic field shifts associated with the magnetic susceptibility of deoxyhemoglobin can be used to estimate SvO₂ , a measure of oxygen extraction and metabolism. However, the distinct magnetic susceptibility of fat surrounding targeted veins will give rise to magnetic field perturbations that will extend into the vein and surrounding tissues, potentially confounding the calculation of SvO₂ .

METHODS

Multi-echo modified Dixon fat-water separated imaging was used to quantify fat-water distributions around the superficial femoral vein (venous return from the lower leg). Fat fraction images were used to generate χ_Fat images, to calculate and remove the associated fat-susceptibility-induced magnetic field shifts before the estimation of SvO₂ . This approach was evaluated at rest and with plantar flexion exercise to evaluate calf muscle oxygen extraction in 10 healthy subjects.

RESULTS

The presence of fat around the vein resulted in complex magnetic field shifts and errors in estimated SvO₂ . Corrected resting SvO₂ values were significantly larger than those measured with conventional methods, at rest (72.6 ± 11.0% vs. 65.2 ± 12.2%, P < 0.05) and post-exercise (37.4 ± 12.3% vs. 31.7 ± 12.7%, P < 0.05), with larger errors in individuals and/or regions with increased fat volumes. Estimation and removal of the field-effects from χ_Fat enabled the use of fat tissues for the measurement and removal of the background magnetic field.

CONCLUSIONS

The magnetic susceptibility effects of fat can confound SvO₂ estimation, but the susceptibility field effects can estimated and removed with the use of modified Dixon fat-water separated imaging.

IDEAL-IQ in an oncologic population: meeting the challenge of concomitant liver fat and liver iron

Background

Cancer patients often have a history of chemotherapy, putting them at increased risk of liver toxicity and pancytopenia, leading to elevated liver fat and elevated liver iron respectively. T1-in-and-out-of-phase, the conventional MR technique for liver fat assessment, fails to detect elevated liver fat in the presence of concomitantly elevated liver iron. IDEAL-IQ is a more recently introduced MR fat quantification method that corrects for multiple confounding factors, including elevated liver iron.

Methods

This retrospective study was approved by the institutional review board with a waiver for informed consent. We reviewed the MRI studies of 50 cancer patients (30 males, 20 females, 50–78 years old) whose exams included (1) T1-in-and-out-of-phase, (2) IDEAL-IQ, and (3) T2* mapping. Two readers independently assessed fat and iron content from conventional and IDEAL-IQ MR methods. Intraclass correlation coefficient (ICC) was estimated to evaluate agreement between conventional MRI and IDEAL-IQ in measuring R2* level (a surrogate for iron level), and in measuring fat level. Agreement between the two readers was also assessed. Wilcoxon signed rank test was employed to compare iron level and fat fraction between conventional MRI and IDEAL-IQ.

Results

Twenty percent of patients had both elevated liver iron and moderate/severe hepatic steatosis. Across all patients, there was high agreement between readers for IDEAL-IQ fat fraction (ICC = 0.957) and IDEAL R2* (ICC = 0.971) measurements, but lower agreement for conventional fat fraction measurements (ICC = 0.626). The fat fractions calculated with IOP were statistically significantly different from those calculated with IDEAL-IQ (reader 1: p < 0.001, reader 2: p < 0.001).

Conclusion

Fat measurements using IDEAL-IQ and IOP diverged in patients with concomitantly elevated liver fat and liver iron. Given prior work validating IDEAL-IQ, these diverging measurements indicate that IOP is inadequate to screen for hepatic steatosis in our cancer population.

Accuracy and precision of proton density fat fraction measurement across field strengths and scan intervals: A phantom and human study

BACKGROUND

Complex-based chemical shift imaging-based magnetic resonance imaging (CSE-MRI) is emerging as a preferred method for noninvasively quantifying proton density fat fraction (PDFF), a promising quantitative imaging biomarker (QIB) for longitudinal hepatic steatosis measurement.

PURPOSE

To determine linearity, bias, repeatability, and reproducibility of the PDFF measurement using CSE-MRI (CSE-PDFF) across scan intervals, MR field strengths, and readers in phantom and nonalcoholic fatty liver disease (NAFLD) patients.

STUDY TYPE

Institutional Review Board (IRB)-approved prospective.

SUBJECTS

Fat-water phantom and 20 adult patients.

FIELD STRENGTH/SEQUENCE

1.5 T and 3.0 T MR systems and a commercially available CSE-MRI sequence (IDEAL-IQ).

ASSESSMENT

Two independent readers measured CSE-PDFF of fat-water phantom and NAFLD patients across two field strengths and scan intervals (same-day and 2-week) each and in a combination of both. MR spectroscopy-based PDFF (MRS-PDFF) was used as the reference standard for phantom PDFF.

STATISTICAL TESTS

Linearity and bias of measurement were evaluated by linear regression analysis and Bland-Altman plots, respectively. Repeatability and reproducibility were assessed by coefficient of variance and repeatability / reproducibility coefficients (RC). The intraclass correlation coefficient was used to validate intra- and interobserver agreements.

RESULTS

CSE-PDFF showed high linearity and small bias (-0.6-0.4 PDFF%) with 95% limits of agreement within ±2.9 PDFF% across field strengths, 2-week interscan period, and readers in the clinical scans. CSE-PDFF was highly repeatable and reproducible both in phantom and clinical scans, with the largest observed RC across field strengths and 2-week interscan period being 3 PDFF%.

DATA CONCLUSION

CSE-PDFF is a robust QIB with high linearity, small bias, and excellent repeatability/reproducibility. A change of more than 3 PDFF% across field strengths within 2 weeks of scan interval likely reflects a true change, which is well within the clinically acceptable range.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:305-314.

Comparison of CT and magnetic resonance mDIXON-Quant sequence in the diagnosis of mild hepatic steatosis

OBJECTIVE:

To determine the diagnostic performance of CT in the assessment of mild hepatic steatosis by comparison with MR mDIXON-Quant as a reference standard, and to explore their clinical applications.

METHODS:

In this prospective study 169 volunteers were included. Each subject underwent CT and MR mDIXON-Quant examinations. Hepatic steatosis evaluations were performed via liver attenuation alone (CT L), liver to spleen attenuation ratio (CT L/S), difference between liver and spleen attenuation (CT L-S), and MR mDIXON-Quant imaging. The effectiveness of CT L, CT L/S, and CT L-S in diagnosing hepatic steatosis severity of ≥5%, ≥10%, and ≥15% was compared, using mDIXON-Quant results as standard.

RESULTS:

65 subjects exhibited mild hepatic steatosis. Hepatic steatosis measurement with mDIXON-Quant was strongly correlated with the three CT methods. Using cutoff value, the sensitivity and specificity of diagnosing hepatic steatosis ≥5, ≥10, and ≥15% were 64.6, 91.3, 100%, and 90.4, 89.7, 93.0% for CT L; 50.8, 87.0, 100%, and 96.2, 98.6, 97.5% for CT L/S; and 67.7, 87.0, 100%, and 81.7, 98.6, 97.5% for CT L-S, respectively. ROC analysis indicated that 58.9, 56.5, and 52.8 HU for CT L; 1.06, 0.98, and 0.90 HU for CT L/S; and 6.21,-1.04, and -4.93 HU for CT L-S were cutoff values for diagnosing hepatic steatosis ≥5%,≥10%, and ≥15%, respectively.

CONCLUSIONS:

The three CT methods exhibit better agreements with mDIXON-Quant imaging for diagnosing hepatic steatosis ≥10%. Hence, CT and mDIXON-Quant could serve as suitable tools for the accurate quantification of mild hepatic steatosis.

SIGNIFICANT FINDS OF THE STUDY:

The close agreement between the three different CT methods (based on our cutoff values) and mDIXON-Quant imaging suggests that CT could accurately diagnose hepatic steatosis ≥10%. Thus, CT and mDIXON-Quant imaging can accurately measure mild hepatic steatosis.

WHAT THIS STUDY ADDS:

Only few studies have compared hepatic steatosis quantification between CT and mDIXON-Quant. We are the first to determine the diagnostic performance of unenhanced CT for quantitatively assessing mild hepatic steatosis, in reference to magnetic resonance mDIXON-Quant imaging.

Assessment of MR-based R2* and quantitative susceptibility mapping for the quantification of liver iron concentration in a mouse model at 7T

PURPOSE

To assess the feasibility of quantifying liver iron concentration (LIC) using R2* and quantitative susceptibility mapping (QSM) at a high field strength of 7 Tesla (T).

METHODS

Five different concentrations of Fe-dextran were injected into 12 mice to produce various degrees of liver iron overload. After mice were sacrificed, blood and liver samples were harvested. Ferritin enzyme-linked immunosorbent assay (ELISA) and inductively coupled plasma mass spectrometry were performed to quantify serum ferritin concentration and LIC. Multiecho gradient echo MRI was conducted to estimate R2* and the magnetic susceptibility of each liver sample through complex nonlinear least squares fitting and a morphology enabled dipole inversion method, respectively.

RESULTS

Average estimates of serum ferritin concentration, LIC, R2*, and susceptibility all show good linear correlations with injected Fe-dextran concentration; however, the standard deviations in the estimates of R2* and susceptibility increase with injected Fe-dextran concentration. Both R2* and susceptibility measurements also show good linear correlations with LIC (R2  = 0.78 and R2  = 0.91, respectively), and a susceptibility-to-LIC conversion factor of 0.829 ppm/(mg/g wet) is derived.

CONCLUSION

The feasibility of quantifying LIC using MR-based  R2* and QSM at a high field strength of 7T is demonstrated. Susceptibility quantification, which is an intrinsic property of tissues and benefits from being field-strength independent, is more robust than R2* quantification in this ex vivo study. A susceptibility-to-LIC conversion factor is presented that agrees relatively well with previously published QSM derived results obtained at 1.5T and 3T.

Effect of microscopic susceptibility gradients on chemical-shift-based fat fraction quantification in supraclavicular fat

BACKGROUND

Susceptibility differences between fat and water can cause changes in the water-fat frequency separation that can negatively affect the accuracy of fat fraction techniques. This may be especially relevant for brown adipose tissue, as MRI fat fraction techniques have been proposed for its detection.

PURPOSE

To assess the effect of microscopic magnetic susceptibility gradients on the water-fat frequency separation and its impact on chemical-shift-based fat fraction quantification techniques in the supraclavicular fat, where brown adipose tissue is commonly found in humans.

STUDY TYPE

Prospective.

POPULATION/SUBJECTS/PHANTOM/SPECIMEN/ANIMAL MODEL

Subjects: 11 healthy volunteers, mean age of 26 and mean BMI of 23, three overweight volunteers, mean age of 38 and mean BMI of 33. Phantoms: bovine phantom and intralipid fat emulsion. Simulations: various water-fat distributions.

FIELD STRENGTH/SEQUENCE

Six-echo gradient echo chemical-shift-encoded sequence at 3T.

ASSESSMENT

Fat fraction values as obtained from a water-fat spectral model accounting for susceptibility-induced water-fat frequency variations were directly compared to traditional spectral models that assume constant water-fat frequency separation.

STATISTICAL TESTS

Two-tail t-tests were used for significance testing (p < 0.05.) A Bayesian Information Criterion difference of 6 between fits was taken as strong evidence of an improved model.

RESULTS

Phantom experiments and simulation results showed variations of the water-fat frequency separation up to 0.4 ppm and 0.6 ppm, respectively. In the supraclavicular area, the water-fat frequency separation produced by magnetic susceptibility gradients varied by as much as ±0.4 ppm, with a mean of 0.08 ± 0.14 ppm, producing a mean difference in fat fraction of -1.26 ± 5.26%.

DATA CONCLUSION

In the supraclavicular fat depot, microscopic susceptibility gradients that exist within a voxel between water and fat compartments can produce variations in the water-fat frequency separation. These variations may produce fat fraction quantification errors of 5% when a spectral model with a fixed water-fat frequency separation is applied, which could impact MR brown fat techniques.

LEVEL OF EVIDENCE

1 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;49:141-151.

Time-resolved contrast-enhanced MR angiography with single-echo Dixon fat suppression

PURPOSE

Dixon-based fat suppression has recently gained interest for dynamic contrast-enhanced MRI, but multi-echo techniques require longer scan times and reduce temporal resolution compared to single-echo alternatives without fat suppression. The purpose of this work is to demonstrate accelerated single-echo Dixon imaging with high spatial and temporal resolution.

THEORY AND METHODS

Real-valued water and fat images can be obtained from a single measurement if the shared initial phase and that due to ΔB0 are assumed known a priori. An expression for simultaneous sensitivity encoding (SENSE) unfolding and fat-water separation is derived for the general undersampling case, and simplified under the special case of uniform Cartesian undersampling. In vivo experiments were performed in extremities and brain with SENSE acceleration factors of up to R = 8.

RESULTS

Single-echo Dixon reconstruction of highly undersampled data was successfully demonstrated. Dynamic contrast-enhanced water and fat images provided high spatial and temporal resolution dynamic images with image update times shorter than previous single-echo Dixon work.

CONCLUSION

Time-resolved contrast-enhanced MRI with single-echo Dixon fat suppression shows high image quality, improved vessel delineation, and reduced sensitivity to motion when compared to time-subtraction methods.

On the sensitivity of quantitative susceptibility mapping for measuring trabecular bone density

Purpose

To develop a methodological framework to simultaneously measure R2* and magnetic susceptibility in trabecularized yellow bone marrow and to investigate the sensitivity of Quantitative Susceptibility Mapping (QSM) for measuring trabecular bone density using a non‐UTE multi‐gradient echo sequence.

Methods

The ankle of 16 healthy volunteers and two patients was scanned using a time‐interleaved multi‐gradient‐echo (TIMGRE) sequence. After field mapping based on water–fat separation methods and background field removal based on the Laplacian boundary value method, three different QSM dipole inversion schemes were implemented. Mean susceptibility values in regions of different trabecular bone density in the calcaneus were compared to the corresponding values in the R2* maps, bone volume to total volume ratios (BV/TV) estimated from high resolution imaging (in 14 subjects), and CT attenuation (in two subjects). In addition, numerical simulations were performed in a simplified trabecular bone model of randomly positioned spherical bone inclusions to verify and compare the scaling of R2* and susceptibility with BV/TV.

Results

Differences in calcaneus trabecularization were well depicted in susceptibility maps, in good agreement with high‐resolution MR and CT images. Simulations and in vivo scans showed a linear relationship of measured susceptibility with BV/TV and R2*. The ankle in vivo results showed a strong linear correlation between susceptibility and R2* (R² = 0.88, p < 0.001) with a slope and intercept of −0.004 and 0.2 ppm, respectively.

Conclusions

A method for multi‐paramteric mapping, including R2*‐mapping and QSM was developed for measuring trabecularized yellow bone marrow, showing good sensitivity of QSM for measuring trabecular bone density.

Validation of a standardized MRI method for liver fat and T2* quantification

Purpose

Several studies have demonstrated the accuracy, precision, and reproducibility of proton density fat fraction (PDFF) quantification using vendor-specific image acquisition protocols and PDFF estimation methods. The purpose of this work is to validate a confounder-corrected, cross-vendor, cross field-strength, in-house variant LMS IDEAL of the IDEAL method licensed from the University of Wisconsin, which has been developed for routine clinical use.

Methods

LMS IDEAL is implemented using a combination of patented and/or published acquisition and some novel model fitting methods required to correct confounds which result from the imaging and estimation processes, including: water-fat ambiguity; T2* relaxation; multi-peak fat modelling; main field inhomogeneity; T1 and noise bias; bipolar readout gradients; and eddy currents. LMS IDEAL has been designed to use image acquisition protocols that can be installed on most MRI scanners and cloud-based image processing to provide fast, standardized clinical results. Publicly available phantom data were used to validate LMS IDEAL PDFF calculations against results from originally published IDEAL methodology. LMS PDFF and T2* measurements were also compared with an independent technique in human volunteer data (n = 179) acquired as part of the UK Biobank study.

Results

We demonstrate excellent agreement of LMS IDEAL across vendors, field strengths, and over a wide range of PDFF and T2* values in the phantom study. The performance of LMS IDEAL was then assessed in vivo against widely accepted PDFF and T2* estimation methods (LMS Dixon and LMS T2*, respectively), demonstrating the robustness of LMS IDEAL to potential sources of error.

Conclusion

The development and clinical validation of the LMS IDEAL algorithm as a chemical shift-encoded MRI method for PDFF and T2* estimation contributes towards robust, unbiased applications for quantification of hepatic steatosis and iron overload, which are key features of chronic liver disease.

Quantitative evaluation of T2* relaxation times for the differentiation of acute benign and malignant vertebral body fractures

OBJECTIVES

The aim of this prospective study was to evaluate the diagnostic performance of T2*-weighted magnetic resonance imaging (MRI) to differentiate between acute benign and neoplastic vertebral compression fractures (VCFs).

MATERIALS AND METHODS

Thirty-seven consecutive patients with a total of 52 VCFs were prospectively enrolled in this IRB approved study. All VCFs were categorized as either benign or malignant according to direct bone biopsy and histopathologic confirmation. In addition to routine clinical spine MRI including at least sagittal T1-weighted, T2-weighted and T2 spectral attenuated inversion recovery (SPAIR)-weighted sequences, all patients underwent an additional sagittal six-echo modified Dixon gradient-echo sequence of the spine at 3.0-T. Intravertebral T2* and T2*_ratio (fracture T2*/normal vertebrae T2*) for acute benign and malignant VCFs were calculated using region-of-interest analysis and compared between both groups. Additional receiver operating characteristic analyses were performed. Five healthy subjects were scanned three times to determine the short-term reproducibility of vertebral T2* measurements.

RESULTS

There were 27 acute benign and 25 malignant VCFs. Both T2* and T2*_ratio of malignant VCFs were significantly higher compared to acute benign VCFs (T2*, 30 ± 11 vs. 19 ± 11 ms [p = 0.001]; T2*_ratio, 2.9 ± 1.6 vs. 1.2 ± 0.7 [p < 0.001]). The areas under the curve were 0.77 for T2* and 0.88 for T2*_ratio, yielding an accuracy of 73% and 89% for distinguishing acute benign from malignant VCFs. The root mean square absolute precision error was 0.44 ms as a measure for the T2* short-term reproducibility.

CONCLUSION

Quantitative assessment of vertebral bone marrow T2* relaxation times provides good diagnostic accuracy for the differentiation of acute benign and malignant VCFs.

Fat-Water Phantoms for Magnetic Resonance Imaging Validation: A Flexible and Scalable Protocol

As new techniques are developed to image adipose tissue, methods to validate such protocols are becoming increasingly important. Phantoms, experimental replicas of a tissue or organ of interest, provide a low cost, flexible solution. However, without access to expensive and specialized equipment, constructing stable phantoms with high fat fractions (e.g., >50% fat fraction levels such as those seen in brown adipose tissue) can be difficult due to the hydrophobic nature of lipids. This work presents a detailed, low cost protocol for creating 5x 100 mL phantoms with fat fractions of 0%, 25%, 50%, 75%, and 100% using basic lab supplies (hotplate, beakers, etc.) and easily accessible components (distilled water, agar, water-soluble surfactant, sodium benzoate, gadolinium-diethylenetriaminepentacetate (DTPA) contrast agent, peanut oil, and oil-soluble surfactant). The protocol was designed to be flexible; it can be used to create phantoms with different fat fractions and a wide range of volumes. Phantoms created with this technique were evaluated in the feasibility study that compared the fat fraction values from fat-water magnetic resonance imaging to the target values in the constructed phantoms. This study yielded a concordance correlation coefficient of 0.998 (95% confidence interval: 0.972-1.00). In summary, these studies demonstrate the utility of fat phantoms for validating adipose tissue imaging techniques across a range of clinically relevant tissues and organs.

MRI adipose tissue and muscle composition analysis-a review of automation techniques

MRI is becoming more frequently used in studies involving measurements of adipose tissue and volume and composition of skeletal muscles. The large amount of data generated by MRI calls for automated analysis methods. This review article presents a summary of automated and semi-automated techniques published between 2013 and 2017. Technical aspects and clinical applications for MRI-based adipose tissue and muscle composition analysis are discussed based on recently published studies. The conclusion is that very few clinical studies have used highly automated analysis methods, despite the rapidly increasing use of MRI for body composition analysis. Possible reasons for this are that the availability of highly automated methods has been limited for non-imaging experts, and also that there is a limited number of studies investigating the reproducibility of automated methods for MRI-based body composition analysis.

3D R 2 * mapping of the placenta during early gestation using free-breathing multiecho stack-of-radial MRI at 3T

BACKGROUND

Multiecho gradient-echo Cartesian MRI characterizes placental oxygenation by quantifying R 2 * . Previous research was performed at 1.5T using breath-held 2D imaging during later gestational age (GA).

PURPOSE

To evaluate the accuracy and repeatability of a free-breathing (FB) 3D multiecho gradient-echo stack-of-radial technique (radial) for placental R 2 * mapping at 3T and report placental R 2 * during early GA.

STUDY TYPE

Prospective.

POPULATION

Thirty subjects with normal pregnancies and three subjects with ischemic placental disease (IPD) were scanned twice: between 14-18 and 19-23 weeks GA.

FIELD STRENGTH

3T.

SEQUENCE

FB radial.

ASSESSMENT

Linear correlation (concordance coefficient, ρ_c ) and Bland-Altman analyses (mean difference, MD) were performed to evaluate radial R 2 * mapping accuracy compared to Cartesian in a phantom. Radial R 2 * mapping repeatability was characterized using the coefficient of repeatability (CR) between back-to-back scans. The mean and spatial coefficient of variation (CV) of R 2 * was determined for all subjects, and separately for anterior and posterior placentas, at each GA range.

STATISTICAL TESTS

ρ_c was tested for significance. Differences in mean R 2 * and CV were tested using Wilcoxon Signed-Rank and Rank-Sum tests. P < 0.05 was considered significant. Z-scores for the IPD subjects were determined.

RESULTS

FB radial demonstrated accurate (ρ_c ≥0.996; P < 0.001; |MD|<0.2s^-1 ) and repeatable (CR<4s^-1 ) R 2 * mapping in a phantom, and repeatable (CR≤4.6s^-1 ) R 2 * mapping in normal subjects. At 3T, placental R 2 * mean ± standard deviation was 12.9s^-1  ± 2.7s^-1 for 14-18 and 13.2s^-1  ± 1.9s^-1 for 19-23 weeks GA. The CV was significantly greater (P = 0.043) at 14-18 (0.63 ± 0.12) than 19-23 (0.58 ± 0.13) weeks GA. At 19-23 weeks, the CV was significantly lower (P < 0.001) for anterior (0.49 ± 0.08) than posterior (0.67 ± 0.11) placentas. One IPD subject had a lower mean R 2 * than normal subjects at both GA ranges (Z<-2).

DATA CONCLUSION

FB radial provides accurate and repeatable 3D R 2 * mapping for the entire placenta at 3T during early GA.

LEVEL OF EVIDENCE

2 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;49:291-303.

Multiparametric quantitative MRI assessment of thigh muscles in limb-girdle muscular dystrophy 2A and 2B

INTRODUCTION

The aim of this study was to apply quantitative MRI (qMRI) to assess structural modifications in thigh muscles of subjects with limb girdle muscular dystrophy (LGMD) 2A and 2B with long disease duration.

METHODS

Eleven LGMD2A, 9 LGMD2B patients and 11 healthy controls underwent a multi-parametric 3T MRI examination of the thigh. The protocol included structural T1-weighted images, DIXON sequences for fat fraction calculation, T2 values quantification and diffusion MRI. Region of interest analysis was performed on 4 different compartments (anterior compartment, posterior compartment, gracilis, sartorius).

RESULTS

Patients showed high levels of fat infiltration as measured by DIXON sequences. Sartorius and anterior compartment were more infiltrated in LGMD2B than LGMD2A patients. T2 values were mildly reduced in both disorders. Correlations between clinical scores and qMRI were found.

CONCLUSIONS

qMRI measures may help to quantify muscular degeneration, but careful interpretation is needed when fat infiltration is massive. Muscle Nerve 58: 550-558, 2018.

Noise properties of proton density fat fraction estimated using chemical shift-encoded MRI

PURPOSE

The purpose of this work is to characterize the noise distribution of proton density fat fraction (PDFF) measured using chemical shift-encoded MRI, and to provide alternative strategies to reduce bias in PDFF estimation.

THEORY

We derived the probability density function for PDFF estimated using chemical shift-encoded MRI, and found it to exhibit an asymmetric noise distribution that contributes to signal-to-noise-ratio dependent bias.

METHODS

To study PDFF noise bias, we performed (at 1.5 T) numerical simulations, phantom acquisitions, and a retrospective in vivo experiment. In each experiment, we compared the performance of three statistics (mean, median, and maximum likelihood estimator) in estimating the PDFF in a region of interest.

RESULTS

We demonstrated the presence of the asymmetric noise distribution in simulations, phantoms, and in vivo. In each experiment we demonstrated that both the median and proposed maximum likelihood estimator statistics outperformed the mean statistic in mitigating noise-related bias for low signal-to-noise-ratio acquisitions.

CONCLUSIONS

Characterization of the noise distribution of PDFF estimated using chemical shift-encoded MRI enabled new strategies based on median and maximum likelihood estimator statistics to mitigate noise-related bias for accurate PDFF measurement from a region of interest. Such strategies are important for quantitative chemical shift-encoded MRI applications that typically operate in low signal-to-noise-ratio regimes. Magn Reson Med 80:685-695, 2018. © 2018 International Society for Magnetic Resonance in Medicine.

Noninvasive, Quantitative Assessment of Liver Fat by MRI-PDFF as an Endpoint in NASH Trials

Nonalcoholic fatty liver disease (NAFLD) is currently the most common cause of chronic liver disease worldwide, and the progressive form of this condition, nonalcoholic steatohepatitis (NASH), has become one of the leading indications for liver transplantation. Despite intensive investigations, there are currently no United States Food and Drug Administration-approved therapies for treating NASH. A major barrier for drug development in NASH is that treatment response assessment continues to require liver biopsy, which is invasive and interpreted subjectively. Therefore, there is a major unmet need for developing noninvasive, objective, and quantitative biomarkers for diagnosis and assessment of treatment response. Emerging data support the use of magnetic resonance imaging-derived proton density fat fraction (MRI-PDFF) as a noninvasive, quantitative, and accurate measure of liver fat content to assess treatment response in early-phase NASH trials. In this review, we discuss the role and utility, including potential sample size reduction, of MRI-PDFF as a quantitative and noninvasive imaging-based biomarker in early-phase NASH trials. Nonalcoholic fatty liver disease (NAFLD) is currently the most common cause of chronic liver disease worldwide.⁽⁾ NAFLD can be broadly classified into two categories: nonalcoholic fatty liver, which has a minimal risk of progression to cirrhosis, and nonalcoholic steatohepatitis (NASH), the more progressive form of NAFLD, which has a significantly increased risk of progression to cirrhosis.⁽⁾ Over the past two decades, NASH-related cirrhosis has become the second leading indication for liver transplantation in the United States.⁽⁾ For these reasons, pharmacological therapy for NASH is needed urgently. Despite intensive investigations, there are currently no therapies for treating NASH that have been approved by the United States Food and Drug Administration.⁽⁾.

Fat Quantification in the Abdomen

Fatty liver disease is characterized histologically by hepatic steatosis, the abnormal accumulation of lipid in hepatocytes. It is classified into alcoholic fatty liver disease and nonalcoholic fatty liver disease, and is an increasingly important cause of chronic liver disease and cirrhosis. Assessing the severity of hepatic steatosis in these conditions is important for diagnostic and prognostic purposes, as hepatic steatosis is potentially reversible if diagnosed early. The criterion standard for assessing hepatic steatosis is liver biopsy, which is limited by sampling error, its invasive nature, and associated morbidity. As such, noninvasive imaging-based methods of assessing hepatic steatosis are needed. Ultrasound and computed tomography are able to suggest the presence of hepatic steatosis based on imaging features, but are unable to accurately quantify hepatic fat content. Since Dixon's seminal work in 1984, magnetic resonance imaging has been used to compute the signal fat fraction from chemical shift-encoded imaging, commonly implemented as out-of-phase and in-phase imaging. However, signal fat fraction is confounded by several factors that limit its accuracy and reproducibility. Recently, advanced chemical shift-encoded magnetic resonance imaging methods have been developed that address these confounders and are able to measure the proton density fat fraction, a standardized, accurate, and reproducible biomarker of fat content. The use of these methods in the liver, as well as in other abdominal organs such as the pancreas, adrenal glands, and adipose tissue will be discussed in this review.

Quantification of Liver Fat Content With Unenhanced MDCT: Phantom and Clinical Correlation With MRI Proton Density Fat Fraction

OBJECTIVE

The purpose of this study was to evaluate the relation between unenhanced CT liver attenuation values and MRI-derived proton density fat fraction (PDFF) for estimation of liver fat content at CT.

MATERIALS AND METHODS

A CT-MRI phantom was constructed and imaged containing 12 vials with lipid fractions ranging from 0% to 100%. For the retrospective clinical arm, 221 patients (120 men, 101 women; mean age, 54 years) underwent both unenhanced CT and chemical shift-encoded MRI of the liver between 2007 and 2017. Among these patients, 92 had more than one 120-kV CT scan for comparison. CT attenuation and MRI PDFF were derived with coregistered ROI measurements in the right hepatic lobe. The 120-kV subgroup of CT examinations performed within 1 month of MRI PDFF examinations (n = 72) served as the primary cohort for linear correlation. The effects of different tube voltage settings, time intervals between CT and MRI, and iron overload were assessed. Linear least squares regression analysis was performed.

RESULTS

Phantom results showed excellent linear fit between CT attenuation and MRI PDFF (r² = 0.986). In patients, 120-kV CT performed within 1 month of MRI PDFF exhibited strong linear correlation (r² = 0.828) that closely matched the phantom data, yielding the following clinical CT-MRI conversion formula: MRI PDFF (%) = -0.58 × CT attenuation (HU) + 38.2. Correlation worsened for CT-to-MRI intervals longer than 1 month (r² = 0.565), and this specific relationship did not apply as well to non-120-kV settings (r² = 0.554). For patients with multiple scans, correlation progressively worsened over time. CT-based liver fat content was underestimated in several patients with iron overload.

CONCLUSION

The linear correlation between unenhanced CT attenuation and MRI PDFF allows quantification of liver fat content by means of unenhanced CT in clinical practice. As expected, correlation worsened with increasing CT-MRI time interval, variable tube voltage settings, and iron overload.

Brown adipose tissue in young adults who were born preterm or small for gestational age

BACKGROUND

Brown adipose tissue (BAT) is present and functions to dissipate energy as heat in young adults and can be assessed using magnetic resonance imaging (MRI) to estimate the voxel fat fraction, i.e. proton density fat fraction (PDFF). It is hypothesized that subjects born preterm or small for gestational age (SGA) may exhibit disrupted BAT formation coupled to metabolic factors. Our purpose was to assess the presence of BAT in young adults born extremely preterm or SGA in comparison with controls.

METHODS

We studied 30 healthy subjects (median age, 21 years): 10 born extremely preterm, 10 full term but SGA and 10 full term with a normal birth weight (controls). We utilized an MRI technique combining multiple scans to enable smaller echo spacing and an advanced fat-water separation method applying graph cuts to estimate B0 inhomogeneity. We measured supraclavicular/cervical PDFF, R2*, fat volume, insulin-like growth factor 1, glucagon, thyroid stimulating hormone and the BAT-associated hormones fibroblast growth factor 21 and irisin.

RESULTS

The groups did not significantly differ in supraclavicular/cervical PDFF, R2*, fat volume or hormone levels. The mean supraclavicular/cervical PDFF was equivalent between the groups (range 75-77%).

CONCLUSIONS

Young adults born extremely preterm or SGA show BAT development similar to those born full term at a normal birth weight. Thus, the increased risk of cardiovascular and metabolic disorders in these groups is not due to the absence of BAT, although our results do not exclude possible BAT involvement in this scenario. Larger studies are needed to understand these relationships.

Fat spectral modeling on triglyceride composition quantification using chemical shift encoded magnetic resonance imaging

PURPOSE

To explore, at a high field strength of 7T, the performance of various fat spectral models on the quantification of triglyceride composition and proton density fat fraction (PDFF) using chemical-shift encoded MRI (CSE-MRI).

METHODS

MR data was acquired from CSE-MRI experiments for various fatty materials, including oil and butter samples and in vivo brown and white adipose mouse tissues. Triglyceride composition and PDFF were estimated using various a priori 6- or 9-peak fat spectral models. To serve as references, NMR spectroscopy experiments were conducted to obtain material specific fat spectral models and triglyceride composition estimates for the same fatty materials. Results obtained using the spectroscopy derived material specific models were compared to results obtained using various published fat spectral models.

RESULTS

Using a 6-peak fat spectral model to quantify triglyceride composition may lead to large biases at high field strengths. When using a 9-peak model, triglyceride composition estimations vary greatly depending on the relative amplitudes of the chosen a priori spectral model, while PDFF estimations show small variations across spectral models. Material specific spectroscopy derived spectral models produce estimations that better correlate with NMR spectroscopy estimations in comparison to those obtained using non-material specific models.

CONCLUSION

At a high field strength of 7T, a material specific 9-peak fat spectral model, opposed to a widely accepted or generic human liver model, is necessary to accurately quantify triglyceride composition when using CSE-MRI estimation methods that assume the spectral model to be known as a priori information. CSE-MRI allows for the quantification of the spatial distribution of triglyceride composition for certain in vivo applications. Additionally, PDFF quantification is shown to be independent of the chosen a priori spectral model, which agrees with previously reported results obtained at lower field strengths (e.g. 3T).

Iron deposition quantification: Applications in the brain and liver

Iron has long been implicated in many neurological and other organ diseases. It is known that over and above the normal increases in iron with age, in certain diseases there is an excessive iron accumulation in the brain and liver. MRI is a noninvasive means by which to image the various structures in the brain in three dimensions and quantify iron over the volume of the object of interest. The quantification of iron can provide information about the severity of iron-related diseases as well as quantify changes in iron for patient follow-up and treatment monitoring. This article provides an overview of current MRI-based methods for iron quantification, specifically for the brain and liver, including: signal intensity ratio, R₂ , R2*, R2', phase, susceptibility weighted imaging and quantitative susceptibility mapping (QSM). Although there are numerous approaches to measuring iron, R₂ and R2* are currently preferred methods in imaging the liver and QSM has become the preferred approach for imaging iron in the brain.

LEVEL OF EVIDENCE

5 Technical Efficacy: Stage 5 J. Magn. Reson. Imaging 2018. J. MAGN. RESON. IMAGING 2018;48:301-317.

Z-spectrum appearance and interpretation in the presence of fat: Influence of acquisition parameters

PURPOSE

Chemical exchange saturation transfer (CEST) MRI is increasingly evolving from brain to body applications. One of the known problems in the body imaging is the presence of strong lipid signals. Although their influence on the CEST effect is acknowledged, there was no study that focuses on the interplay among echo time, fat fraction, and Z-spectrum. This study strives to address these points, with the emphasis on the application in the breast.

METHODS

Z-spectra were simulated in phase and out of phase of the main fat peak at -3.4 ppm, with the fat fraction varying from 0 to 100%. The magnetization transfer ratio asymmetry in two ranges, centering at the exchanging pool and at 3.5 ppm approximately opposite the nonexchanging fat pool, were calculated and were plotted against fat fraction. The results were verified in phantoms and in vivo.

RESULTS

The results demonstrate the combined influence of fat fraction and echo time on the Z-spectrum for gradient echo based CEST acquisitions. The influence is straightforward in the in-phase images, but it is more complicated in the out-of-phase images, potentially leading to erroneous CEST contrast.

CONCLUSIONS

This study provides a basis for understanding the origin and appearance of lipid artifacts in CEST imaging, and lays the foundation for their efficient removal. Magn Reson Med 79:2731-2737, 2018. © 2017 International Society for Magnetic Resonance in Medicine.