Spin-Echo planar spectroscopic imaging for fast lipid characterization in bone marrow

Quantitative 1 H MRI and MRS of fatty acid composition

Adipose tissue as well as other depots of fat (triglycerides) are increasingly being recognized as active contributors to the human function and metabolism. In addition to the fat concentration, also the fatty acid chemical composition (FAC) of the triglyceride molecules may play an important part in diseases such as obesity, insulin resistance, hepatic steatosis, osteoporosis, and cancer. MR spectroscopy and chemical-shift-encoded imaging (CSE-MRI) are established methods for non-invasive quantification of fat concentration in tissue. More recently, similar techniques have been developed for assessment also of the FAC in terms of the number of double bonds, the fraction of saturated, monounsaturated, and polyunsaturated fatty acids, or semi-quantitative unsaturation indices. The number of papers focusing on especially CSE-MRI-based techniques has steadily increased during the past few years, introducing a range of acquisition protocols and reconstruction algorithms. However, a number of potential sources of bias have also been identified. Furthermore, the measures used to characterize the FAC using both MRI and MRS differ, making comparisons between different techniques difficult. The aim of this paper is to review MRS- and MRI-based methods for in vivo quantification of the FAC. We describe the chemical composition of triglycerides and discuss various potential FAC measures. Furthermore, we review acquisition and reconstruction methodology and finally, some existing and potential applications are summarized. We conclude that both MRI and MRS provide feasible non-invasive alternatives to the gold standard gas chromatography for in vivo measurements of the FAC. Although both are associated with gas chromatography, future studies are warranted.

Fat-water separation by fast metabolite cycling magnetic resonance spectroscopic imaging at 3 T: A method to generate separate quantitative distribution maps of musculoskeletal lipid components

PURPOSE

To provide a rapid, noninvasive fat-water separation technique that allows producing quantitative maps of particular lipid components.

METHODS

The calf muscles in 5 healthy adolescents (age 12-16 years; body mass index = 20 ± 3 kg/m² ) were scanned by two different fat fraction measurement methods. A density-weighted concentric-ring trajectory metabolite-cycling MRSI technique was implemented to collect data with a nominal resolution of 0.25 mL within 3 minutes and 16 seconds. For comparative purposes, the standard Dixon technique was performed. The two techniques were compared using structural similarity analysis. Additionally, the difference in the distribution of each lipid over the adolescent calf muscles was assessed based on the MRSI data.

RESULTS

The proposed MRSI technique provided individual fat fraction maps for eight musculoskeletal lipid components identified by LCModel analysis (IMC/L [CH₃ ], EMCL [CH₃ ], IMC/L [CH₂ ]_n , EMC/L [CH₂ ]_n , IMC/L [CH₂ -CH], EMC/L [CH₂ -CH], IMC/L [-CH=CH-], and EMC/L [-CH=CH-]) with mean structural similarity indices of 0.19, 0.04, 0.03, 0.50, 0.45, 0.04, 0.07, and 0.12, respectively, compared with the maps generated by the used Dixon method. Further analysis of voxels with zero structural similarity demonstrated an increased sensitivity of fat fraction lipid maps from the data acquired using this MRSI technique over the standard Dixon technique. The lipid spatial distribution over calf muscles was consistent with previously published findings in adults.

CONCLUSION

This MRSI technique can be a useful tool when individual lipid fat fraction maps are desired within a clinically acceptable time and with a nominal spatial resolution of 0.25 mL.

Quantitative assessment of bone marrow edema-like lesion and overlying cartilage in knees with osteoarthritis and anterior cruciate ligament tear using MR imaging and spectroscopic imaging at 3 Tesla

PURPOSE

To quantitatively assess bone marrow edema-like lesion (BMEL) and the overlying cartilage in osteoarthritis (OA) or anterior cruciate ligament (ACL)-injured knees using magnetic resonance imaging (MRI) and spectroscopic imaging (MRSI).

MATERIALS AND METHODS

Eight healthy controls and 30 patients with OA and other injuries who showed BMEL were scanned at 3.0T. A regression model was constructed to automatically calculate the volume of BMEL. Normalized T(1rho) z-scores were calculated within BMEL-overlying cartilage. Three-dimensional (3D) MRSI was acquired in the BMEL and surrounding bone marrow.

RESULTS

The mean T(1rho) z-score was significantly higher in BMEL-overlying cartilage than that in surrounding cartilage in the lateral tibia of patients with ACL tears (P < 0.001). Significantly elevated water and unsaturated lipids, and decreased saturated lipids were observed in BMEL. The volume of elevated water correlated with the volume of BMEL. Water content was significantly higher within BMEL than that outside BMEL. The unsaturation index outside BMEL was significantly higher in patients with ACL tears than that in OA.

CONCLUSION

3D MRSI and T(1rho) mapping provide tools to quantitatively evaluate BMEL in OA and knee injury. This may allow us to better understand pathophysiology and evolution of injuries and degenerative conditions of the knee.

Temperature monitoring with line scan echo planar spectroscopic imaging

A new magnetic resonance imaging method, line scan echo planar spectroscopic imaging (LSEPSI), is shown capable of providing rapid, internally referenced temperature monitoring from water and fat chemical shifts.

METHODS

Orthogonal 90 degrees and 180 degrees slice selective RF pulses inclined by 45 degrees from the image plane solicit a spin echo from a tissue column. The echo is read by asymmetric sampling of 32 gradient echoes spaced 1.4-1.8 ms apart. Sixty-four adjacent columns are sequentially sampled in 4.2-6.4 s with 4,096 voxels sampled with voxel volumes of 0.08-0.13 cm3. Mixed mayonnaise/water phantoms were used to correlate LSEPSI-derived chemical shifts and thermocouple-based temperature measurements from 23 to 60 degrees C with a 1.5 T scanner. Measurement artifacts unrelated to temperature were investigated with the phantom, as was the feasibility of applying the sequence in human breast in vivo.

RESULTS

The correlation between LSEPSI and thermocouple-based temperature measurements in the phantom was excellent (r2>0.99). Field drifts affecting the temperature measurements using the water peak alone were corrected by using the water/lipid peak difference. The sequence had an average temperature resolution of 1.4 degrees C in the phantom. The frequency difference measurement reduced the sensitivity to artifacts related to temperature. Both water and lipid peaks were detectable throughout many locations in the breast, suggesting the applicability of LSEPSI in this organ.

DISCUSSION

T1-saturation losses occur in conventional and echo-planar based 2D CSI sequences using phase encoding methods with short TR periods. These losses are eliminated when individual columns are sampled in snapshot fashion with LSEPSI since the effective TR becomes the time between scans rather than excitations. T1 saturation can make small spectral peaks difficult to detect at high temperatures and generally lowers the signal-to-noise ratio of the spectra. The rapid acquisition and insensitivity to T1 saturation effects make LSEPSI an attractive technique for monitoring thermal therapies in breast using the internally referenced fat/water frequency separation.

Continuous distribution analysis of marrow 1 H magnetic resonance relaxation in bone

Longitudinal and transverse NMR relaxation of 1H nuclei were studied in vitro on fresh animal femur samples. A large number of data points were taken, starting at 100 micros for T(1) by inversion-recovery, at 200 micros for T(2) by single-echo sequences, and at 600 micros for T(2) by CPMG echo-trains. Quasi-continuous distributions of relaxation times were computed, giving wide distributions for all samples. Bulk marrow removed from the medullary cavity showed T(2) distributions from about 20 ms to 600 ms and T(1) distributions from about 40 ms to 2 s. The 1H nuclei in trabecular bone samples, where marrow is confined, may show long tails for T(2) at relaxation times down to 250 micros, the origin of which is still not known. These tails are absent in bulk marrow from the medullary cavity. The differences observed in T(1) distributions among trabecular bone samples are in accordance with the different marrow compositions. Discrete exponential fits were computed also, and in most cases four discrete exponential components were required to fit the experimental data adequately. However, the discrete components do not seem to correspond to any physically distinguishable separate compartments.

Line scan echo planar spectroscopic imaging

A line scan echo planar spectroscopic imaging (LSEPSI) sequence is presented which can rapidly produce 2D chemical shift imaging (CSI) data with minimal relaxation weighting and motion-related artifacts. The technique is based on successive "snapshot" 1D CSI acquisitions of individual tissue columns, and avoids T(1) saturation problems associated with the short TR periods needed for very rapid scanning with either conventional or echo planar-based 2D CSI methods. Potential applications include rapid fat/water spectral quantitation in the abdomen and internally referenced temperature monitoring for interventional procedures.

High-speed spectroscopic imaging for cancellous bone marrow R(2)* mapping and lipid quantification

In this work an interleaved multiple-gradient-echo chemical shift imaging (IMGE-CSI) technique was designed, implemented and evaluated at 1.5 and 4T for high-resolution lipid quantification and R(2)* measurement in-vivo. The method is analogous to echo planar CSI but utilizes conventional gradient echoes, exploiting the principle of spectroscopic bandwidth extension by interleaving temporally offset gradient-echo trains. It is shown that IMGE-CSI is able to measure true fat volume fraction in oil/water mixtures with high accuracy, not possible with Dixon-type methods which approximate the spectrum as consisting of only two spectral components. Correlation of the CSI- derived volume fractions with volumetry afforded r(2) > 0.99 with a slope of 0.98. The method is shown to be able to quantify regional variations in bone marrow composition in vivo with a spatial resolution of 2.5 x 2.5 x 5 mm(3.) R(2)* was obtained by multi-line spectral curve fitting. For the measurement of R(2)* in cancellous bone marrow the method is shown to agree well with time-domain fitting techniques but is superior in instances where the marrow has both hematopoietic and fatty constituents. Finally, excellent inter-scan reproducibility (1% coefficient of variation for global means and medians) was achieved, yielding r(2) = 0.98 of the test-retest correlation for three scans in four test subjects. In conclusion, IMGE-CSI is found to enable highly accurate lipid quantification and measurement of cancellous bone marrow R(2)* at spatial resolutions and scan times typical of standard clinical protocols.