Limits of Fat Quantification in the Presence of Iron Overload.

BACKGROUND: Chemical shift encoded magnetic resonance imaging (CSE-MRI)-based tissue fat quantification is confounded by increased R2* signal decay rate caused by the presence of excess iron deposition. PURPOSE: To determine the upper limit of R2* above which it is no longer feasible to quantify proton density fat fraction (PDFF) reliably, using CSE-MRI. STUDY TYPE: Prospective. POPULATION: Cramér-Rao lower bound (CRLB) calculations, Monte Carlo simulations, phantom experiments, and a prospective study in 26 patients with known or suspected liver iron overload. FIELD STRENGTH/SEQUENCE: Multiecho gradient echo at 1.5 T and 3.0 T. ASSESSMENT: CRLB calculations were used to develop an empirical relationship between the maximum R2* value above which PDFF estimation will achieve a desired number of effective signal averages. A single voxel multi-TR, multi-TE stimulated echo acquisition mode magnetic resonance spectroscopy acquisition was used as a reference standard to estimate PDFF. Reconstructed PDFF and R2* maps were analyzed by one analyst using multiple regions of interest drawn in all nine Couinaud segments. STATISTICAL TESTS: None. RESULTS: Simulations, phantom experiments, and in vivo measurements demonstrated unreliable PDFF estimates with increased R2*, with PDFF errors as large as 20% at an R2* of 1000 s-1 . For typical optimized Cartesian acquisitions (TE1 = 0.75 msec, ΔTE = 0.67 msec at 1.5 T, TE1 = 0.65 msec, ΔTE = 0.58 msec at 3.0 T), an empirical relationship between PDFF estimation errors and acquisition parameters was developed that suggests PDFF estimates are unreliable above an R2* of ~538 s-1 and ~779 s-1 at 1.5 T and 3 T, respectively. This empirical relationship was further investigated with phantom experiments and in vivo measurements, with PDFF errors at an R2* of 1000 s-1 at 3.0 T as large as 10% with TE1 = 1.24 msec, ΔTE = 1.01 msec compared to 3% with TE1 = 0.65 msec, ΔTE = 0.58 msec. DATA CONCLUSION: We successfully developed a theoretically-based empirical formula that may provide an easily calculable guideline to identify R2* values above which PDFF is not reliable in research and clinical applications using CSE-MRI to quantify PDFF in the presence of iron overload. LEVEL OF EVIDENCE: 1 TECHNICAL EFFICACY STAGE: 1.

Accuracies of Chemical Shift In/Opposed Phase and Chemical Shift Encoded Magnetic Resonance Imaging to Detect Intratumoral Fat in Hepatocellular Carcinoma.

BACKGROUND: Magnetic Resonance Imaging (MRI) being a noninvasive modality may help in preoperative evaluation of intratumoral fat in hepatocellular carcinoma (HCC) using chemical shift encoded (CSE) MRI and in-/opposed-phase (IOP) imaging sequences. PURPOSE: To compare the diagnostic accuracy of chemical shift encoded fat fraction at three different flip angles (FAs) using quantitative chemical shift encoded MRI (CSE-MRI) with in-/opposed phase (IOP) imaging to evaluate intratumoral fat in HCC. STUDY TYPE: Retrospective. POPULATION: Eighty-six patients with 87 pathology proven HCCs. FIELD STRENGTH/SEQUENCE: IOP (LAVA-Flex) and CSE-MRI (IDEAL IQ) a three-dimensional spoiled gradient-echo pulse sequences acquired at 3 T. ASSESSMENT: Regions of interest (ROIs) were manually drawn by two observers in the tumors to measure mean fat fractions. Surgical specimens were reassessed for intratumoral fat content. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were assessed for CSE-MRI sequence at FA 3°, 8°, and 9°. STATISTICAL TESTS: Intraclass correlation coefficient (ICC) was expressed in terms of inter- and intra-observer agreements. Receiver operating characteristic curve analysis was performed for the diagnostic performance followed by combined metric of both. SNR/CNR were analyzed by Kruskal-Wallis test. RESULTS: Excellent inter- and intra-observer agreements (ICC >0.95, P < 0.001) were observed for both IOP and CSE-MRI. IOP (86.4%) showed higher sensitivity than CSE-MRI at FA 3° (72.5%), FA 8° (76.4%) and FA 9° (76.3%). In contrast, the specificity for CSE-MRI at FA 3° (86%), FA 8° (87%), and FA 9° (87%) were greater than IOP (72%). A combined metric of IOP and CSE-MRI derived fat fractions at FA 8° gave highest AUC of 87% and accuracy of 86%. SNR and CNR for CSE-MRI were significantly higher at FA 8° and FA 9° than FA 3° (P < 0.05). DATA CONCLUSION: IOP and quantitative CSE-MRI are both feasible methods to detect intratumoral fat in HCC with higher accuracy and SNR for CSE-MRI at FA 8° and 9°. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 2.

Water-fat magnetic resonance imaging of adipose tissue compartments in the normal third trimester fetus.

BACKGROUND: Assessment of fetal adipose tissue gives information about the future metabolic health of an individual, with evidence that the development of this tissue has regional heterogeneity. OBJECTIVE: To assess differences in the proton density fat fraction (PDFF) between fetal adipose tissue compartments in the third trimester using water-fat magnetic resonance imaging (MRI). MATERIALS AND METHODS: Water-fat MRI was performed in a 1.5-T scanner. Fetal adipose tissue was segmented into cheeks, thorax, abdomen, upper arms, forearms, thighs and lower legs. PDFF and R2* values were measured in each compartment. RESULTS: Twenty-eight women with singleton pregnancies were imaged between 28 and 38 weeks of gestation. At 30 weeks' gestation (n=22), the PDFF was statistically different between the compartments (P<0.0001), with the highest PDFF in cheeks, followed by upper arms, thorax, thighs, forearms, lower legs and abdomen. There were no statistical differences in the rate of PDFF change with gestational age between the white adipose tissue compartments (P=0.97). Perirenal brown adipose tissue had a different PDFF and R2* compared to white adipose tissue, while the rate of R2* change did not significantly change with gestational age between white adipose tissue compartments (P=0.96). CONCLUSION: Fetal adipose tissue accumulates lipids at a similar rate in all white adipose tissue compartments. PDFF variances between the compartments suggest that accumulation begins at different gestational ages, starting with cheeks, followed by extremities, trunk and abdomen. Additionally, MRI was able to detect differences in the PDFF between fetal brown adipose tissue and white adipose tissue.

Validation of goose liver fat measurement by QCT and CSE-MRI with biochemical extraction and pathology as reference.

OBJECTIVES: This study aimed to validate the accuracy and reliability of quantitative computed tomography (QCT) and chemical shift encoded magnetic resonance imaging (CSE-MRI) to assess hepatic steatosis. METHODS: Twenty-two geese with a wide range of hepatic steatosis were collected. After QCT and CSE-MRI examinations, the liver of each goose was removed and samples were taken from the left lobe, upper and lower half of the right lobe for biochemical measurement and histology. Fat percentages by QCT and proton density fat fraction by MRI (MRI-PDFF) were measured within the sample regions of biochemical measurement and histology. The accuracy of QCT and MR measurements were assessed through Spearman correlation coefficients (r) and Passing and Bablok regression equations using biochemical measurement as the "gold standard". RESULTS: Both QCT and MRI correlated highly with chemical extraction [r = 0.922 (p < 0.001) and r = 0.949 (p < 0.001) respectively]. Chemically extracted triglyceride was accurately predicted by both QCT liver fat percentages (Y = 0.6 + 0.866 × X) and by MRI-PDFF (Y = -1.8 + 0.773 × X). CONCLUSIONS: QCT and CSE-MRI measurements of goose liver fat were accurate and reliable compared with biochemical measurement. KEY POINTS: • QCT and CSE-MRI can measure liver fat content accurately and reliably • Histological grading of hepatic steatosis has larger sampling variability • QCT and CSE-MRI have potential in the clinical setting.

Quantification of liver fat in the presence of iron overload.

PURPOSE: To evaluate the accuracy of R2* models (1/T2 * = R2*) for chemical shift-encoded magnetic resonance imaging (CSE-MRI)-based proton density fat-fraction (PDFF) quantification in patients with fatty liver and iron overload, using MR spectroscopy (MRS) as the reference standard. MATERIALS AND METHODS: Two Monte Carlo simulations were implemented to compare the root-mean-squared-error (RMSE) performance of single-R2* and dual-R2* correction in a theoretical liver environment with high iron. Fatty liver was defined as hepatic PDFF >5.6% based on MRS; only subjects with fatty liver were considered for analyses involving fat. From a group of 40 patients with known/suspected iron overload, nine patients were identified at 1.5T, and 13 at 3.0T with fatty liver. MRS linewidth measurements were used to estimate R2* values for water and fat peaks. PDFF was measured from CSE-MRI data using single-R2* and dual-R2* correction with magnitude and complex fitting. RESULTS: Spectroscopy-based R2* analysis demonstrated that the R2* of water and fat remain close in value, both increasing as iron overload increases: linear regression between R2*W and R2*F resulted in slope = 0.95 [0.79-1.12] (95% limits of agreement) at 1.5T and slope = 0.76 [0.49-1.03] at 3.0T. MRI-PDFF using dual-R2* correction had severe artifacts. MRI-PDFF using single-R2* correction had good agreement with MRS-PDFF: Bland-Altman analysis resulted in -0.7% (bias) ± 2.9% (95% limits of agreement) for magnitude-fit and -1.3% ± 4.3% for complex-fit at 1.5T, and -1.5% ± 8.4% for magnitude-fit and -2.2% ± 9.6% for complex-fit at 3.0T. CONCLUSION: Single-R2* modeling enables accurate PDFF quantification, even in patients with iron overload. LEVEL OF EVIDENCE: 1 J. Magn. Reson. Imaging 2017;45:428-439.

Improved fat-suppression homogeneity with mDIXON turbo spin echo (TSE) in pediatric spine imaging at 3.0 T.

Background Robust fat suppression remains essential in clinical MRI to improve tissue signal contrast, minimize fat-related artifacts, and enhance image quality. Purpose To compare fat suppression between mDIXON turbo spin echo (TSE) and conventional frequency-selective and inversion-recovery methods in pediatric spine MRI. Material and Methods Images from T1-weighted (T1W) and T2-weighted (T2W) TSE sequences coupled with conventional methods and the mDIXON technique were compared in 36 patients (5.8 ± 5.4 years) at 3.0 T. Images from 42 pairs of T1W (n = 16) and T2W (n = 26) scans were acquired. Two radiologists reviewed the data and rated images using a three-point scale in two categories, including the uniformity of fat suppression and overall diagnostic image quality. The Wilcoxon rank-sum test was used to compare the scores. Results The Cohen's kappa coefficient for inter-rater agreement was 0.69 (95% confidence interval [CI], 0.56-0.83). Images from mDIXON TSE were considered superior in fat suppression ( P < 0.01) in 22 (rater 1) and 25 (rater 2) cases, respectively. In 13 (rater 1) and 11 (rater 2) cases, mDIXON TSE demonstrated improved diagnostic image quality ( P < 0.01). In three cases, fat suppression was superior using inversion-recovery and likewise in one case mDIXON had poorer image diagnostic quality. Lastly, mDIXON and conventional fat-suppression methods performed similarly in 17 (rater 1) and 14 (rater 2) cases, and yielded equal diagnostic image quality in 28 (rater 1) and 30 (rater 2) cases. Conclusion Robust fat suppression can be achieved with mDixon TSE pediatric spine imaging at 3.0 T and should be considered as a permanent replacement of traditional methods, in particular frequency-selective techniques.

Comparison of the value of adipose tissues in abdomen and lumbar vertebra for predicting disease activity in Crohn's disease: A preliminary study based on CSE-MRI.

BACKGROUND: To compare the value of adipose tissues in abdomen and lumbar vertebra for predicting Crohn's disease (CD) activity based on chemical shift encoded magnetic resonance imaging (CSE-MRI). METHODS: 84 CD patients were divided into remission, mild, and moderate-severely groups based on CD activity index (CDAI). Differences in different adipose parameters [subcutaneous adipose tissue (SAT), visceral adipose tissue (VAT), mesenteric fat index (MFI), and bone marrow fat fraction (BMFF)] and blood inflammatory indicators among three groups, as well as the correlation of above parameters and CDAI were analyzed. The areas under the receiver-operating characteristic curves (AUCs) for the parameters selected by multivariate logistic regression analysis for predicting active CD were calculated. RESULTS: There were no significant differences in VAT and MFI among three groups (both P > 0.05). The cross-sectional areas of SAT in moderate-severe group were significantly lower than those in remission group (P = 0.014). BMFF values of remission group were significantly higher than those in the mild and moderate-severe groups (both P < 0.001). BMFF was negatively correlated with CDAI (r = -0.595, P < 0.001). SAT exhibited no significant correlation with CDAI. Erythrocyte sedimentation rate (ESR) and BMFF were the independent predictors of CDAI. Both combined had a higher diagnostic efficacy for active CD with an AUC of 0.895. CONCLUSIONS: BMFF is the best marker for predicting CD activity in fat parameters of abdomen and lumbar vertebra based on CSE-MRI. The model based on BMFF and ESR has a high efficiency in predicting active CD. TRIAL REGISTRATION: No. 22 K164 (Registered 18-07-2022).

Semiautomated segmentation of lower extremity MRI reveals distinctive subcutaneous adipose tissue in lipedema: a pilot study.

Purpose: Lipedema is a painful subcutaneous adipose tissue (SAT) disease involving disproportionate SAT accumulation in the lower extremities that is frequently misdiagnosed as obesity. We developed a semiautomatic segmentation pipeline to quantify the unique lower-extremity SAT quantity in lipedema from multislice chemical-shift-encoded (CSE) magnetic resonance imaging (MRI). Approach: Patients with lipedema (n=15) and controls (n=13) matched for age and body mass index (BMI) underwent CSE-MRI acquired from the thighs to ankles. Images were segmented to partition SAT and skeletal muscle with a semiautomated algorithm incorporating classical image processing techniques (thresholding, active contours, Boolean operations, and morphological operations). The Dice similarity coefficient (DSC) was computed for SAT and muscle automated versus ground truth segmentations in the calf and thigh. SAT and muscle volumes and the SAT-to-muscle volume ratio were calculated across slices for decades containing 10% of total slices per participant. The effect size was calculated, and Mann-Whitney U test applied to compare metrics in each decade between groups (significance: two-sided P<0.05). Results: Mean DSC for SAT segmentations was 0.96 in the calf and 0.98 in the thigh, and for muscle was 0.97 in the calf and 0.97 in the thigh. In all decades, mean SAT volume was significantly elevated in participants with versus without lipedema (P<0.01), whereas muscle volume did not differ. Mean SAT-to-muscle volume ratio was significantly elevated (P<0.001) in all decades, where the greatest effect size for distinguishing lipedema was in the seventh decade approximately midthigh (r=0.76). Conclusions: The semiautomated segmentation of lower-extremity SAT and muscle from CSE-MRI could enable fast multislice analysis of SAT deposition throughout the legs relevant to distinguishing patients with lipedema from females with similar BMI but without SAT disease.

Correction of QCT vBMD using MRI measurements of marrow adipose tissue.

OBJECTIVE: Quantitative computed tomography (QCT) measurements of volumetric bone mineral density (vBMD) are subject to errors due to variations in the amount of bone marrow adipose tissue (BMAT). The purpose of our study was to describe and validate a novel method to correct lumbar spine trabecular vBMD measurements for BMAT using chemical shift-encoded magnetic resonance imaging (CSE-MRI). METHODS: CSE-MRI measurements of proton density fat fraction (PDFF) were used to correct QCT spine vBMD measurements for BMAT based on the H2O and K2HPO4 basis set equivalent densities of bone, red and yellow bone marrow. BMAT corrected and uncorrected vBMD measurements of the L1 vertebra were compared with dual-energy QCT (DEQCT) measurements in 18 subjects (mean age: 68 y, range 60 to 93 y). A further 400 subjects (mean age: 53 y, range 21 to 82 y) had 120 kVp single-energy QCT and CES-MRI scans of L2-L4 and the data used to simplify the adipose tissue correction by deriving a linear equation between the CSE-MRI vBMD correction and fractional BMAT content. RESULTS: Application of the CSE-MRI derived vBMD correction changed the bias (95% limits of agreement) compared with DEQCT from 26.7 (11.0 to 42.4) mg/cm3 to 2.2 (-9.5 to 13.9) mg/cm3 at 80 kVp, and from 22.4 (3.3 to 41.6) mg/cm3 to 2.9 (-12.6 to 18.4) mg/cm3 at 120 kVp. Data for the 400 subjects gave the following relationship valid at 120 kVp: vBMD correction (mg/cm3) = -12.96 + 75.76 × BMAT. CONCLUSION: CSE-MRI measurements of PDFF can be used to correct for BMAT content and improve the accuracy of lumbar spine QCT vBMD measurements calibrated using a K2HPO4 phantom.

High SNR Acquisitions Improve the Repeatability of Liver Fat Quantification Using Confounder-corrected Chemical Shift-encoded MR Imaging.

PURPOSE: To determine whether high signal-to-noise ratio (SNR) acquisitions improve the repeatability of liver proton density fat fraction (PDFF) measurements using confounder-corrected chemical shift-encoded magnetic resonance (MR) imaging (CSE-MRI). MATERIALS AND METHODS: Eleven fat-water phantoms were scanned with 8 different protocols with varying SNR. After repositioning the phantoms, the same scans were repeated to evaluate the test-retest repeatability. Next, an in vivo study was performed with 20 volunteers and 28 patients scheduled for liver magnetic resonance imaging (MRI). Two CSE-MRI protocols with standard- and high-SNR were repeated to assess test-retest repeatability. MR spectroscopy (MRS)-based PDFF was acquired as a standard of reference. The standard deviation (SD) of the difference (Δ) of PDFF measured in the two repeated scans was defined to ascertain repeatability. The correlation between PDFF of CSE-MRI and MRS was calculated to assess accuracy. The SD of Δ and correlation coefficients of the two protocols (standard- and high-SNR) were compared using F-test and t-test, respectively. Two reconstruction algorithms (complex-based and magnitude-based) were used for both the phantom and in vivo experiments. RESULTS: The phantom study demonstrated that higher SNR improved the repeatability for both complex- and magnitude-based reconstruction. Similarly, the in vivo study demonstrated that the repeatability of the high-SNR protocol (SD of Δ = 0.53 for complex- and = 0.85 for magnitude-based fit) was significantly higher than using the standard-SNR protocol (0.77 for complex, P < 0.001; and 0.94 for magnitude-based fit, P = 0.003). No significant difference was observed in the accuracy between standard- and high-SNR protocols. CONCLUSION: Higher SNR improves the repeatability of fat quantification using confounder-corrected CSE-MRI.