Spin locking for magnetic resonance imaging with application to human breast

A pulse sequence for rapid in vivo spin-locked MRI

PURPOSE

To develop a novel pulse sequence called spin-locked echo planar imaging (EPI), or (SLEPI), to perform rapid T1rho-weighted MRI.

MATERIALS AND METHODS

SLEPI images were used to calculate T1rho maps in two healthy volunteers imaged on a 1.5-T Sonata Siemens MRI scanner. The head and extremity coils were used for imaging the brain and blood in the popliteal artery, respectively.

RESULTS

SLEPI-measured T1rho was 83 msec and 103 msec in white (WM) and gray matter (GM), respectively, 584 msec in cerebrospinal fluid (CSF), and was similar to values obtained with the less time-efficient sequence based on a turbo spin-echo readout. T1rho was 183 msec in arterial blood at a spin-lock (SL) amplitude of 500 Hz.

CONCLUSION

We demonstrate the feasibility of the SLEPI pulse sequence to perform rapid T1rho MRI. The sequence produced images of higher quality than a gradient-echo EPI sequence for the same contrast evolution times. We also discuss applications and limitations of the pulse sequence.

T1ρ Dispersion profile of rat tissues in vitro at very low locking fields

The purpose of this study was to show the T(1rho) dispersion profile in various rat tissues (liver, brain, spleen, kidney, heart and skeletal muscle) at low (0.1 T) B(0) field at very low locking field B1, starting from 10 microT. The T(1rho) dispersion profile showed a quite similar pattern in all tissues. The highest R(1rho) relaxation rates were seen in the liver and muscle followed by the heart, whereas the values for spleen, kidney and brain were rather similar. The greatest difference between R2 relaxation rate and R(1rho) relaxation rate at B1=10 microT was seen in the liver and muscle. The steepest slope for a dispersion curve was seen in the muscle. The value of T(1rho) approximately approached the value of T2 when the locking field B1 approached 0. Except for the liver, the calculated apparent relaxation rate R2' was slightly larger than the calculated one. The potential value of T(1rho) imaging is to combine high R1 contrast of low-field imaging with the high signal-to-noise ratio (SNR) of high static field imaging. T(1rho) relaxation and dispersion data presented in the current study help to optimize the rotating-frame MR imaging.

Quantitative T(1rho) and adiabatic Carr-Purcell T2 magnetic resonance imaging of human occipital lobe at 4 T

The feasibility of performing quantitative T(1rho) MRI in human brain at 4 T is shown. T(1rho) values obtained from five volunteers were compared with T2 and adiabatic Carr-Purcell (CP) T2 values. Measured relaxation time constants increased in order from T2, CP-T2, T(1rho) both in white and gray matter, demonstrating differential sensitivities of these methods to dipolar interactions and/or proton exchange and diffusion in local microscopic field gradients, which are so-called dynamic averaging (DA) processes. In occipital lobe, all relaxation time constants were found to be higher in white matter than in gray matter, demonstrating contrast denoted as an "inverse transverse relaxation contrast." This contrast persisted despite changing the delay between refocusing pulses or changing the magnitude of the spin-lock field strength, which suggests that it does not originate from DA, as might be induced by the presence of Fe, but rather is related to dipolar interactions in the brain tissue.

T2rho-weighted contrast in MR images of the human brain

In this work, the feasibility of using T2rho weighting as an MR contrast mechanism is evaluated. Axial images of a human brain were acquired using a single-slice spin-lock T2rho-weighted pulse sequence and compared to analogous T2-weighted images of the same slice. The contrast between white matter and gray matter in T2rho-weighted images was approximately 40% greater than that from T2-weighted data. These preliminary data suggest that the novel contrast mechanism of T2rho can be used to yield high-contrast T2-like images.

T1, T2 relaxation and magnetization transfer in tissue at 3T

T1, T2, and magnetization transfer (MT) measurements were performed in vitro at 3 T and 37 degrees C on a variety of tissues: mouse liver, muscle, and heart; rat spinal cord and kidney; bovine optic nerve, cartilage, and white and gray matter; and human blood. The MR parameters were compared to those at 1.5 T. As expected, the T2 relaxation time constants and quantitative MT parameters (MT exchange rate, R, macromolecular pool fraction, M0B, and macromolecular T2 relaxation time, T2B) at 3 T were similar to those at 1.5 T. The T1 relaxation time values, however, for all measured tissues increased significantly with field strength. Consequently, the phenomenological MT parameter, magnetization transfer ratio, MTR, was lower by approximately 2 to 10%. Collectively, these results provide a useful reference for optimization of pulse sequence parameters for MRI at 3 T.

T1rho-weighted MRI using a surface coil to transmit spin-lock pulses

T1rho-weighted MRI is a novel basis for generating tissue contrast. However, it suffers from sensitivity to B1 inhomogeneity. First, excitation with a spatially varying B1 causes flip-angle artifacts and second, spin locking with an inhomogeneous B1 results in non-uniform T1rho contrast. In this study, we overcome the former complication with a specially designed spin-locking pulse sequence and we successfully obtain T1rho-weighted images with a surface coil. In this pulse sequence, the spin-lock pulse was divided into segments of equal duration and alternating phase. This "self-compensating" T1rho-preparatory pulse sequence was analyzed and the effect of an inhomogeneous B1 field was simulated using the Bloch equations. T1rho-weighted MR images of a phantom and a human knee joint in vivo were obtained on a clinical scanner with a surface coil to demonstrate the utility of the pulse sequence. The self-compensating T1rho-prepared pulses sequence resulted in substantially reduced image artifacts compared to the conventional, single-phase spin-lock pulse.

In vivo quantification ofT1? using a multislice spin-lock pulse sequence

A multislice spin-lock (MS-SL) pulse sequence is implemented on a clinical scanner to acquire multiple images with spin-lock-generated contrast of the knee joints of six healthy human subjects. The MS-SL sequence produces images with T1rho contrast with an additional factor of intrinsic T2rho weighting, which hinders direct measurement of T1rho. A method is presented to compensate the MS-SL-generated data with regard to T2rho in an effort to accurately calculate multislice T1rho maps in a feasible experimental time. The T2rho-compensated multislice T1rho maps produced errors in the measurement of T1rho in healthy patellar cartilage of approximately 5% compared to the gold standard measurement of T1rho acquired with single-slice spin-lock pulse sequence. The MS-SL sequence has potential as an important clinical tool for the acquisition of multislice T1rho-weighted images and/or quantitative multislice T1rho maps.

Pulse sequence for multisliceT1?-weighted MRI

A 2D multislice spin-lock (MS-SL) MR pulse sequence is presented for rapid volumetric T1rho-weighted imaging. Image quality is compared with T1rho-weighted data collected using a single-slice (SS) SL sequence and T2-weighted data from a standard MS spin-echo (SE) sequence. Saturation of longitudinal magnetization by the application of nonselective SL pulses is experimentally measured and theoretically modeled as T2rho decay. The saturation data is used to correct the image data as a function of the SL pulse duration to make quantitative measurements of T1rho. Measurements of T1rho using the saturation-corrected MS-SL data are nearly identical to those measured using an SS-SL sequence. The MS-SL sequence produces quantitative T1rho maps of an entire sample volume with the high-SNR advantages conferred by SE-based sequences.

Method for reduced SART1?-weighted MRI

A reduced specific absorption rate (SAR) version of the T(1rho)-weighted MR pulse sequence was designed and implemented. The reduced SAR method employs a partial k-space acquisition approach in which a full power spin-lock pulse is applied to only the central phase-encode lines of k-space, while the remainder of k-space receives a low-power spin-lock pulse. Acquisition of high- and low-power phase-encode lines are interspersed chronologically to minimize average power deposition. In this way, the majority of signal energy in the central portion of k-space receives full T(1rho)-weighting, while the average SAR of the overall acquisition can be reduced, thereby lowering the minimum safely allowable TR. The pulse sequence was used to create T(1rho) maps of a phantom, an in vivo mouse brain, and the brain of a human volunteer. In the images of the human brain, SAR was reduced by 40% while the measurements of T(1rho) differed by only 2%. The reduced SAR sequence enables T(1rho)-weighted MRI in a clinical setting, even at high field strengths.

Three-dimensional T1rho-weighted MRI at 1.5 Tesla

PURPOSE

To design and implement a magnetic resonance imaging (MRI) pulse sequence capable of performing three-dimensional T(1rho)-weighted MRI on a 1.5-T clinical scanner, and determine the optimal sequence parameters, both theoretically and experimentally, so that the energy deposition by the radiofrequency pulses in the sequence, measured as the specific absorption rate (SAR), does not exceed safety guidelines for imaging human subjects.

MATERIALS AND METHODS

A three-pulse cluster was pre-encoded to a three-dimensional gradient-echo imaging sequence to create a three-dimensional, T(1rho)-weighted MRI pulse sequence. Imaging experiments were performed on a GE clinical scanner with a custom-built knee-coil. We validated the performance of this sequence by imaging articular cartilage of a bovine patella and comparing T(1rho) values measured by this sequence to those obtained with a previously tested two-dimensional imaging sequence. Using a previously developed model for SAR calculation, the imaging parameters were adjusted such that the energy deposition by the radiofrequency pulses in the sequence did not exceed safety guidelines for imaging human subjects. The actual temperature increase due to the sequence was measured in a phantom by a MRI-based temperature mapping technique. Following these experiments, the performance of this sequence was demonstrated in vivo by obtaining T(1rho)-weighted images of the knee joint of a healthy individual.

RESULTS

Calculated T(1rho) of articular cartilage in the specimen was similar for both and three-dimensional and two-dimensional methods (84 +/- 2 msec and 80 +/- 3 msec, respectively). The temperature increase in the phantom resulting from the sequence was 0.015 degrees C, which is well below the established safety guidelines. Images of the human knee joint in vivo demonstrate a clear delineation of cartilage from surrounding tissues.

CONCLUSION

We developed and implemented a three-dimensional T(1rho)-weighted pulse sequence on a 1.5-T clinical scanner.

T1rho MR imaging of the human wrist in vivo

RATIONALE AND OBJECTIVES

The purpose of this study was (a) to demonstrate the feasibility of computing T1rho maps of, and T1rho dispersion in, human wrist cartilage at MR imaging in vivo and (b) to compare T1rho and T2 weighting in terms of magnitude of relaxation times and signal intensity contrast.

MATERIALS AND METHODS

T2 and T1rho magnetic resonance images of wrist joints in healthy volunteers (n = 5) were obtained with a spin-echo sequence and a fast spin-echo sequence pre-encoded with a spin-lock pulse cluster. A 1.5-T clinical imager was used (Signa; GE Medical Systems, Milwaukee, Wis) with a 9.5-cm-diameter transmit-receive quadrature birdcage coil tuned to 63.75 MHz.

RESULTS

T1rho relaxation times at a spin-lock frequency of 500 Hz vary from 40.5 msec +/- 0.85 to 56.6 msec +/- 4.83, and T2 relaxation times vary from 28.1 msec +/- 1.88 to 34.5 msec +/- 2.63 (mean +/- standard error of the mean, n = 5, P < .016) in various regions of the wrist. T1rho dispersion was observed in the range of spin-lock frequencies studied. T1rho-weighted images not only have higher signal-to-noise ratios but also show better fluid and fat signal suppression than T2-weighted images.

CONCLUSION

It was possible to perform T2- and T1rho-weighted MR imaging of human wrist cartilage in vivo with standard clinical imagers. The higher signal-to-noise ratio and improved contrast between cartilage and surrounding fat achieved with T1rho imaging may provide better definition of lesions and accurate quantitation of small changes in cartilage degeneration.

Artifacts in T(1rho)-weighted imaging: correction with a self-compensating spin-locking pulse

Significant artifacts arise in T(1rho)-weighted imaging when nutation angles suffer small deviations from their expected values. These artifacts vary with spin-locking time and amplitude, severely limiting attempts to perform quantitative imaging or measurement of T(1rho) relaxation times. A theoretical model explaining the origin of these artifacts is presented in the context of a T(1rho)-prepared fast spin-echo imaging sequence. Experimentally obtained artifacts are compared to those predicted by theory and related to B(1) inhomogeneity. Finally, a "self-compensating" spin-locking preparatory pulse cluster is presented, in which the second half of the spin-locking pulse is phase-shifted by 180 degrees. Use of this pulse sequence maintains relatively uniform signal intensity despite large variations in flip angle, greatly reducing artifacts in T(1rho)-weighted imaging.

Proton spin-lock ratio imaging for quantitation of glycosaminoglycans in articular cartilage

PURPOSE

To quantify glycosaminoglycans (GAG) in intact bovine patellar cartilage using the proton spin-lock ratio imaging method. This approach exploits spin-lattice relaxation time in the rotating frame (T(1rho)) imaging and T(1rho) relaxivity (R(1rho)).

MATERIALS AND METHODS

All the magnetic resonance imaging (MRI) experiments were performed on a 4-T whole-body GE Signa scanner (GEMS, Milwaukee, WI), and spectroscopy experiments of chondroitin sulfate (CS) phantoms were done on a 2-T custom-built spectrometer. A custom-built 11-cm-diameter transmit-receive birdcage coil, which was tuned to a proton frequency of 170 MHz, was employed for the imaging experiments. T(1rho) measurements were made using a fast spin echo (FSE) sequence pre-encoded with a three-pulse cluster consisting of two 90 degrees hard pulses separated by a low-power rectangle pulse for spin-locking.

RESULTS

The methodology is first validated on CS phantoms and then used to quantify GAG content in intact bovine cartilage (N = 5). There is a good agreement between the GAG map calculated from the T(1rho) ratio imaging method (71 +/- 4%) and GAG measured from spectrophotometric assay (75 +/- 5%) in intact bovine tissue.

CONCLUSION

We have demonstrated a proton spin-lock ratio imaging method to quantify absolute GAG distribution in the cartilage in a noninvasive and nondestructive manner.

On- and off-resonance T(1rho) MRI in acute cerebral ischemia of the rat

The ability of on-resonance T(1rho) (T(1rho)) and off-resonance T(1rho) (T(1rho)(off)) measurements to indicate acute cerebral ischemia in a rat model of transient middle cerebral artery (MCA) occlusion was investigated at 4.7 T. T(1rho) was determined with B(1) fields of 0.4, 0.8, and 1.6 G, and T(1rho)(off) with five offset frequencies ((Delta)omega) ranging from 0-7.5 kHz at B(1) of 0.4 G, yielding effective B(1) (B(eff)) from 0.4 to 1.8 G. Diffusion, T(1), and T(2) were also quantified. Both T(1rho) and T(1rho)(off) acquired with (Delta)(o)< 2.5 kHz showed positive contrast during the first hours of MCA occlusion in the ischemic tissue delineated by low diffusion. Interestingly, T(1rho)(off) contrast acquired with (Delta)omega > 2.5 kHz was clearly less sensitive to ischemic alterations, and developed with a delayed time course. This discrepancy is thought to be a consequence of the frequency dependency of cross-relaxation during irradiation with spin-lock pulses.

[Proton spinlocking and T1 rho-weighted MR imaging at 1.5 T]

During spinlocking, the magnetization is aligned along an oscillating field (RF) and relaxes with time constant T1 rho, the spin-lattice relaxation time in the rotating frame. Using a clinical whole-body MR scanner, methods of spinlocking preparation and signal acquisition were combined to evaluate the potential of T1 rho-weighted MR imaging (T1 rho w-MRI) at B0 = 1.5 T. Examinations of the brain of healthy volunteers yielded images with pronounced contrast and T1 rho-variation of the tissue. However, the contrast resembled that of T2-weighted MRI, which is explained by the restricted spinlocking-field strength (BSL < or = 6 microT) on the tomograph. The result (mono-exponential fit) of serial T1 rho w-MRI data from examinations of 8 volunteers was on average 105 +/- 4 ms in the gray matter and 86 +/- 4 ms in the white matter (for BSL = 3 microT). The values are comparable to T2 of both tissues. MRT with spinlocking is less susceptible to local magnetic field inhomogeneities than conventional MRI.

Equivalent cross-relaxation rate imaging in the synthetic copolymer gels and invasive ductal carcinomas of the breast

The values of equivalent cross-relaxation rate (ECR) correlated well with [i] water conditions in various copolymer gels and [ii] nature of malignant cells with regard to nuclear dysplasia and mitotic potential in breast carcinomas. The synthetic copolymer gels composed of any two or three monomers among 2-hydroxyethyl methacrylate (HEMA), glycidyl methacrylate (GMA), N-vinyl-2-pyrrolidinone (N-VP), methyl methacrylate (MMA) and benzyl methacrylate (BMA). The ECR measurement was performed by using an off-resonance saturation pulse under conventional field-echo imaging at frequency within +/- 75 ppm apart from the water resonance frequency. The ECR values were readily to determine and non-time consuming parameter for cross relaxation rate. The ECR values at the frequency offset by 7-ppm (ECR-7) were divided the sample gels two classes, which must correspond to hydrophilic or hydrophobic ones. The sensitivity in the gels was nearly equivalent to the cross-relaxation rate itself. In the breast carcinomas, the ECR-7 correlates with the nature of malignant cells with regard to nuclear dysplasia and mitotic potential. The ECR-7 is better or more accurate than the STR-7 because the SDNRs between carcinoma and glandular tissue increased by approximately 50% on the ECR-7 compared with the STR-7. Thus the ECR values could be a new parameter for malignancy and cell proliferative activity of the breast carcinomas with non-invasive modalities by magnetic resonance imaging.

Multiple-slice spin lock imaging of head and neck tumors at 0.1 Tesla: exploring appropriate imaging parameters with reference to T2-weighted spin-echo technique

RATIONALE AND OBJECTIVES

Spin lock imaging has been shown to be useful in characterizing head and neck tumors. The purposes of this study were to explore and develop multiple-slice spin lock gradient-echo (SL-GRE) sequences for head and neck imaging and to compare the tumor contrast on SL images to spin-echo (SE) T2-weighted images at 0.1 T.

METHODS

On the basis of measured relaxation times of tumors and head and neck tissues, the authors evaluated with signal equations the effect of imaging parameters on tissue contrast produced by the SL-GRE sequence. In the clinical study, 34 patients with pathologically verified head and neck tumors were imaged with multiple-slice SL-GRE (repetition time 1500 ms/echo time 30 ms) out-of-phase fat/water sequences and compared with T2-weighted SE (repetition time 1500 ms/echo time 120 ms) sequences. The conspicuity of tumors was evaluated by calculating the contrast-to-noise ratios (CNRs).

RESULTS

The combination of a short echo time of 30 ms and the length of locking pulses in the range of 10 to 35 ms produced optimal CNRs for head and neck tumor imaging. The measured CNRs and subjective evaluation for tumor detection were satisfactory with both imaging sequences. However, the CNRs between tumors and salivary gland tissues were significantly greater with the SL sequence than with the T2-weighted sequence.

CONCLUSIONS

The multiple-slice SL-GRE technique provides image contrast comparable to that of SE T2-weighted imaging for head and neck tumors at 0.1 T. With short locking pulse lengths and echo times, wide anatomic coverage and reduced motion and susceptibility artifacts can be achieved. The out-of-phase SL technique is useful in imaging salivary gland tumors.

Proteoglycan-induced changes in T1rho-relaxation of articular cartilage at 4T

Proteoglycan (PG) depletion-induced changes in T1rho (spin-lattice relaxation in rotating frame) relaxation and dispersion in articular cartilage were studied at 4T. Using a spin-lock cluster pre-encoded fast spin echo sequence, T1rho maps of healthy bovine specimens and specimens that were subjected to PG depletion were computed at varying spin-lock frequencies. Sequential PG depletion was induced by trypsinization of cartilage for varying amounts of time. Results demonstrated that over 50% depletion of PG from bovine articular cartilage resulted in average T1rho increases from 110-170 ms. Regression analysis of the data showed a strong correlation (R2 = 0.987) between changes in PG and T1rho. T1rho values were highest at the superficial zone and decreased gradually in the middle zone and again showed an increasing trend in the region near the subchondral bone. The potentials of this method in detecting early degenerative changes of cartilage are discussed. Also, T(1rho)-dispersion changes as a function of PG depletion are described.

Saturation transfer ratio imaging in invasive ductal carcinomas of the breast

A prospective study was performed to investigate the correlations between saturation transfer ratio (STR) and histologic parameters of invasive ductal carcinomas in human breast. The histologic parameters investigated were the extent of fibrosis in the intercellular matrix, dysplastic changes of nuclei, and mitotic index. Twenty-seven patients with breast carcinoma were examined using an off-resonance saturation pulse in conjunction with conventional field-echo T(1)-weighted imaging at frequency offsets of 448 Hz and 1200 Hz from water resonance. The values of STR at frequency offset of 1200 Hz (STR(1200)) increased from non-scirrhous carcinoma to scirrhous carcinoma. Although STR(1200) showed correlation with the extent of fibrosis in the intercellular matrix (p<0.01, n = 27), they did not correlate with the dysplastic changes of nuclei or mitotic index. On the other hand, the values of STR at frequency offset of 448 Hz (STR(448)) demonstrated close correlation to dysplastic changes of nuclei and mitotic index (p<0.01, n = 27). STR(1200) correlates with the structural characteristics and STR(448) correlates with the nature of malignant cells with regard to nuclear dysplasia and mitotic potential.

T1rho imaging using magnetization-prepared projection encoding (MaPPE)

T1rho contrast weighting using a magnetization-prepared projection encoding (MaPPE) pulse sequence was investigated. Fast radial imaging was implemented by applying magnetization preparation pulses, each followed by multiple RF alpha pulses encoding radial trajectories of k-space. Acquiring multiple views per preparatory pulse imposes view-to-view variation; the resultant distortion of the point-spread function is examined. The issue of maximizing signal while preserving the intended contrast weighting is addressed. Under modification of repetition time and flip angle (alpha), three distinct behavior regimes of the sequence are identified. The utility of the pulse sequence as a quantitative relaxation measurement tool is also examined by comparing imaging and spectroscopy experiments. A mouse was imaged in vitro to demonstrate the viability of application to MR histology. These images exhibit the utility of spinlocking and projection encoding as an aftemative contrast source to both T2-weighted MaPPE images and conventional T2-weighted spin-echo images.

T1rho dispersion of rat tissues in vitro

The purpose of this study was to demonstrate T1rho dispersion in different rat tissues (liver, brain, spleen, kidney, heart, and skeletal muscle), and to compare the 1/T1rho data to previous 1/T1 data and magnetization transfer of rat tissues at low (0.1 T) B0 field. The 1/T1rho dispersion showed a fairly similar pattern in all tissues. The highest 1/T1rho relaxation rates were seen in liver and muscle followed by heart, whereas the values for spleen, kidney, and brain were quite similar. Compared to 1/T2 relaxation rate, the greatest difference was seen in liver and muscle. The rank order 1/T1rho value at each locking field B1 was the same as the transfer rate of magnetization from the water to the macromolecular pool (Rwm) for liver, muscle, heart, and brain. The potential value T1rho imaging is to combine high T1 contrast of low field imaging with the high signal to noise ratio of high static field imaging. When the T1rho value for a given tissue is known, the contrast between different tissues can be optimized by adjusting the locking time TL. Further studies are encouraged to fully exploit this. Targets for more detailed research include brain infarct, brain and liver tumors.

3D spin-lock imaging of human gliomas

We investigated whether the simultaneous use of paramagnetic contrast medium and 3D on-resonance spin lock (SL) imaging could improve the contrast of enhancing brain tumors at 0.1 T. A phantom containing serial concentrations of gadopentetate dimeglumine (Gd-DTPA) in cross-linked bovine serum albumin (BSA) was imaged. Eleven patients with histologically verified glioma were also studied. T1-weighted 3D gradient echo images with and without SL pulse were acquired before and after a Gd-DTPA injection. SL effect, contrast, and contrast-to-noise ratio (CNR) were calculated for each patient. In the glioma patients, the SL effect was significantly smaller in the tumor than in the white and gray matter both before (p = 0.001, p = 0.025, respectively), and after contrast medium injection (p < 0.001, p < 0.001, respectively). On post-contrast images, SL imaging significantly improved tumor contrast (p = 0.001) whereas tumor CNR decreased slightly (p = 0.024). The combined use of SL imaging and paramagnetic Gd-DTPA contrast agent offers a modality for improving tumor contrast in magnetic resonance imaging (MRI) of enhancing brain tumors. 3D gradient echo SL imaging has also shown potential to increase tissue characterization properties of MR imaging of human gliomas.

On- and off-resonance spin-lock MR imaging of normal human brain at 0.1 T: possibilities to modify image contrast

The aim of the present investigation was to determine spin lock (SL) relaxation parameters for the normal brain tissues and thus, to provide basis for optimizing the imaging contrast at 0.1 T. 68 healthy volunteers were included. On-resonance spin lock relaxation time (T1rho) and off-resonance spin lock relaxation parameters (T1rho(off), Me/Mo), MT parameters (T1sat, Ms/Mo), and T1, T2 were determined for the cortical gray matter, and for the frontal and parietal white matters. The T1rho for the frontal and parietal white matters ranged from 110 to 133 ms and from 122 to 155 ms with locking field strengths from 50 microT to 250 microT, respectively. Accordingly, the values for the gray matter ranged from 127 to 155 ms. With a locking field strength of 50 microT, T1rho(off) for the frontal and parietal white matters were from 114 to 217 ms and from 126 to 219 ms, and for the gray matter from 136 to 267 ms with the angle between the effective magnetic field (B(eff)) and the z-axis (theta) ranging from 60 degrees to 15 degrees, respectively. The T1rho of the white and gray matters increased significantly with increasing locking field amplitude (p < 0.001). The T1rho(off) decreased significantly with increasing theta (p < 0.001). T1rho and T1rho(off) with theta > or = 30 degrees were statistically significantly shorter in the frontal than in the parietal white matters (p < 0.05). The duration, amplitude and theta of the locking pulse provide additional parameters to optimize contrast in brain SL imaging.

Proton T1rho-dispersion imaging of rodent brain at 1.9 T

Detection of H2(17)O with proton T1rho-dispersion imaging holds promise as a means of quantifying metabolism and blood flow with MRI. However, this technique requires a priori knowledge of the intrinsic T1rho dispersion of tissue. To investigate these properties, we implemented a T1rho imaging sequence on a 1.9-T Signa GE scanner. A series of T1rho images for different locking frequencies and locking durations were obtained from rat brain in vivo and compared with 5% (wt/vol) gelatin phantoms containing different concentrations of (17)O ranging from .037% (natural abundance) to 2.0 atom%. Results revealed that, although there is considerable T1rho-dispersion in phantoms doped with H2(17)O, the T1rho of rat brain undergoes minimal dispersion for spin-locking frequencies between .2 and 1.5 kHz. A small degree of T1rho dispersion is present below .2 kHz, which we postulate arises from natural-abundance H2(17)O. Moreover, the signal-to-noise ratios of T1rho-weighted images are significantly better than comparable T2-weighted images, allowing for improved visualization of tissue contrast. We have also demonstrated the feasibility of proton T1rho-dispersion imaging for detecting intravenous H2(17)O on a live mouse brain. The potential application of this technique to study brain perfusion is discussed.

Spin lock and magnetization transfer MR imaging of local liver lesions

The present study was designed to evaluate tissue contrast characteristics obtained with the spin-lock (SL) technique by comparing the results with those generated with a magnetization transfer(MT)-weighted gradient echo [GRE, echo-time (TE)=40 ms] sequence. Twenty-eight patients with hepatic hemangiomas (n=14), or metastatic liver lesions (n=14) were imaged at 0.1 T by using identical imaging parameters. Gradient echo, single-slice off-resonance MT, and multiple-slice SL sequences were obtained. SL and MT-effects were measured from the focal liver lesions and from normal liver parenchyma. In addition, tissue contrast values for the liver lesions were determined. Statistically significant difference between the SL-effects of the hemangiomas and metastases, and also between the MT-effects of the lesions was observed (p < 0.02). Tissue contrast values for the lesions proved to be quite similar between the SL and MT techniques. Our results indicate that at 0.1 T multiple-slice SL imaging provides MT based tissue contrast characteristics in tissues rich in protein with good imaging efficiency and wide anatomical coverage, and with reduced motion and susceptibility artifacts.

Determination of T1rho values for head and neck tissues at 0.1 T: a comparison to T1 and T2 relaxation times

In order to optimize head and neck magnetic resonance (MR) imaging with the spin-lock (SL) technique, the T1rho relaxation times for normal tissues were determined. Furthermore, T1rho was compared to T1 and T2 relaxation times. Ten healthy volunteers were studied with a 0.1 T clinical MR imager. T1rho values were determined by first measuring the tissue signal intensities with different locking pulse durations (TL), and then by fitting the signal intensity values to the equation with the least-squares method. The T1rho relaxation times were shortest for the muscle and tongue, intermediate for lymphatic and parotid gland tissue and longest for fat. T1rho demonstrated statistically significant differences (p < 0.05) between all tissues, except between muscle and tongue. T1rho values measured at locking field strength (B1L) of 35 microT were close to T2 values, the only exception being fat tissue, which showed T1rho values much longer than T2 values. Determination of tissue relaxation times may be utilized to optimize image contrast, and also to achieve better tissue discrimination potential, by choosing appropriate imaging parameters for the head and neck spin-lock sequences.

Low field T1rho imaging of myositis

The purpose of this study was to evaluate 1/T1rho in relation to 1/T1 and 1/T2 in characterizing normal and diseased muscle. We measured the muscle relaxation rates 1/T1 and 1/T2 at 0.1 T and 1/T1rho at on-resonance locking fields B1 between 10 and 160 microT in myositis patients and normal volunteers. 1/T2 and 1/T1rho of muscle were lower in the patients than in the volunteers, whereas there was no difference in the 1/T1 values. The lower relaxation rates 1/T2 and 1/T1rho in the diseased muscle may be due to fat and connective tissue infiltrations and edema. 1/T1rho contrast between muscle and subcutaneous fat was higher than 1/T2 and 1/T1 contrast. This may be explained by the different B1 dispersion behavior of these two tissue types. 1/T1rho of fat is B1 field independent, whereas 1/T1rho of muscle decreases clearly with increasing B1 field. In conclusion, 1/T1rho provides a useful tool in manipulating contrast in magnetic resonance imaging of diseased muscle.

Off-resonance proton T1rho dispersion imaging of 17O-enriched tissue phantoms

Proton T1rho dispersion imaging is a recently described method for indirect detection of 17O. However, clinical implementation of this technique is hindered by the requirement for a high-amplitude spin-locking field (gammaB1 > 1 kHz) that exceeds current limitations in specific absorption rate (SAR). Here, a strategy is offered for circumventing high SAR in T1rho dispersion imaging of 17O through the use of low-amplitude off-resonance spin-locking pulses (gammaB1 < 300 Hz). Proton spin-lattice relaxation times in the off-resonance rotating frame were measured in H2(17)O-enriched tissue phantoms. On- and off-resonance T1rho dispersion imaging was implemented at 2 T using a spin-locking preparatory pulse cluster appended to a standard spin-echo sequence. On- and off-resonance dispersion images exhibited similar 17O-based image contrast. Magnetization transfer effects did not depend on 17O concentration and had no effect on image contrast. In conclusion, off-resonance proton T1rho dispersion imaging shows promise as a safe, sensitive technique for generating 17O-based T1rho contrast without exceeding SAR limitations.

Comparison of four magnetization preparation schemes to improve blood-wall contrast in cine short-axis cardiac imaging

Improvements in short-axis blood-myocardium contrast in the heart with the use of four magnetization preparation schemes applied before the imaging sequence are demonstrated. Gradient-echo cine cardiac images are acquired and compared at 0.95 T incorporating T2, T1rho, magnetization transfer, and double inversion (black blood) preparations in a series of volunteer studies over the first 550 ms of the cardiac cycle. T2 and T1rho preparations exhibit improvements of 100% and above in image contrast. Magnetization transfer preparation exhibits improvements of 50% in image contrast, whereas an initial improvement (50%) followed by a large loss in contrast is observed using the black blood preparation. Improvements in contrast are dependent on tissue relaxation parameters and therefore are suitable for studies involving patients exhibiting poor in-flow enhancement of blood caused by poor heart function.

T1 rho MR imaging characteristics of human anterior tibial and gastrocnemius muscles

RATIONALE AND OBJECTIVES

The authors evaluated the value of T1 rho in relation to T1 and T2 in the characterization of human muscles.

MATERIALS AND METHODS

The authors studied the effect of muscle type (anterior tibial [AT] and gastrocnemius [GC]), sex, and age on 1/T1 and 1/T2 at 0.1 T and on 1/ T1 rho at locking-field B1s (spin-locking radio-frequency magnetic induction field) of 10-160 microT in 38 healthy volunteers. The contrast-to-noise ratio (CNR) between muscle and fat was evaluated with different T1-, T2-, and T1 rho-weighted magnetic resonance (MR) sequences.

RESULTS

The 1/T1, 1/T2, and 1/T1 rho were slightly higher in AT than in GC muscles. The 1/T2 and 1/T1 rho of AT muscles showed a sex dependence, whereas no correlation with age was found. The CNR of the T1 rho-weighted images did not markedly differ from that of the T1- and T2-weighted images.

CONCLUSION

T1 rho is as sensitive as T2 to the composition of muscle, whereas T1 is less sensitive. In MR imaging of normal muscle, T1 rho and T2 provide a relatively similar tissue contrast.

Magnetic resonance properties of ex vivo breast tissue at 1.5 T

The magnetic resonance absorption spectrum, T1 and T2 relaxation time distributions, and magnetization transfer properties of ex vivo breast tissue have been characterized at 1.5 T and 37 degrees C. The fraction of fibroglandular tissue within individual tissue samples (n = 31) was inferred from the tissue volumetric water content obtained by integration of resolvable broad-line fat and water resonances. The spectroscopically estimated water content was strongly correlated with that extracted enzymatically (Pearson correlation coefficient 0.98, P < < 0.01), which enabled the assignment of principal relaxation components for fibroglandular tissue (T2=0.04+/-0.01, T1=1.33+/-0.24 s), and for adipose tissue (T2=0.13+/-0.01, T1=0.23+/-0.01 s, and T2=0.38+/-0.03, T1=0.62+/-0.16 s). Th e relaxation components for fibroglandular tissue exhibited strong magnetization transfer, whereas those for adipose tissue showed little magnetization transfer effect. These results ultimately have applicability to the optimization of clinical magnetic resonance imaging and research investigations of the breast.

T1 rho dispersion imaging of head and neck tumors: a comparison to spin lock and magnetization transfer techniques

The potential of T1 rho dispersion, spin lock (SL), and magnetization transfer (MT) techniques to differentiate benign and malignant head and neck tumors was evaluated. Twenty-four patients with pathologically verified head and neck tumors were studied with a .1-T MR imager. T1 rho dispersion effect was defined as 1 -(intensity with lower locking field amplitude/intensity with higher locking field amplitude). T1 rho dispersion effects were higher for malignant than benign tumors (P = .001). With T1 rho dispersion effect .14 as the threshold, sensitivity for detecting a malignant tumor was 91%, specificity was 77%, and accuracy was 83%. A strong correlation between T1 rho dispersion effects and SL effects and between T1 rho dispersion effects and MT effects in the head and neck tumors was found (r = .87, P < .001 and r = .90, P < .001, respectively). High T1 rho dispersion effects are not specific indicators of malignancy, because chronic infections, some benign tumors, and malignancies may overlap. Low T1 rho dispersion effect values are characteristic of a benign tumor.

T1rho of protein solutions at very low fields: dependence on molecular weight, concentration, and structure

The effect of molecular weight, concentration, and structure on 1/T1rho, the rotating frame relaxation rate, was investigated for several proteins using the on-resonance spin-lock technique, for locking fields B1 < 200 microT. The measured values of 1/T1rho were fitted to a simple theoretical model to obtain the dispersion curves 1/T1rho(omega1) and the relaxation rate at zero B1 field, 1/T1rho(0). 1/T1rho was highly sensitive to the molecular weight, concentration, and structure of the protein. The amount of intra- and intermolecular hydrogen and disulfide bonds especially contributed to 1/T1rho. In all samples, 1/T1rho(0) was equal to 1/T2 measured at the main magnetic field Bo = 0.1 T, but at higher locking fields the dispersion curves monotonically decreased. The results of this work indicate that a model considering the effective correlation time of molecular motions as the main determinant for T1rho relaxation in protein solutions is not valid at very low B1 fields. The underlying mechanism for the relaxation rate 1/T1rho at B1 fields below 200 microT is discussed.

Application of a quantitative model to differentiate benign from malignant breast lesions detected by dynamic, gadolinium-enhanced MRI

Both benign and malignant breast lesions may exhibit intense contrast enhancement when imaged using gadolinium-enhanced MRI. We propose a quantitative approach for fitting dynamic signal intensity (SI) data that may distinguish benign from malignant lesions. We studied 78 lesions in 75 women (18 malignancies, 16 fibroadenomas, and 44 other benign breast lesions) to determine the potential of this model for decreasing false-positive MR results. Twenty-eight lesions showed no enhancement; all were benign. One lesion showed a complex pattern not amenable to region-of-interest analysis and was considered a false positive. SI versus time data for the remaining 49 lesions were fit to the proposed model. We found that one parameter, M, the normalized slope of the SI enhancement profile evaluated at half the maximal signal intensity, seemed to be highly correlated with malignancy and offered improved discrimination between malignant and benign lesions compared to a previously published two-point slope method.

T1 rho relaxation and its application to MR histology

The application of T1 rho an an alternative contrast parameter in high-field magnetic resonance histology (MRH) has been investigated. Spectroscopic measurements of T1 rho were performed on 5.75% agar and 1.0 mM MnCI2 phantoms at 9.4 T to validate the accuracy of the imaging measurements. Image studies were performed at 2.0 and 9.4 T on perfusion-fixed 17.5-day-old mouse embryos. T1, T2, and T1 rho relaxation times were calculated for the phantoms and muscle, diencephalon, and liver tissues. The 5.75% agar phantom and all tissues showed T1 rho dispersion with B1L, whereas the 1.0 mM MnCI2 phantom showed no significant B1L dependence. T1 rho dispersion with B(O) was observed arising from the effects of diffusion through susceptibility-induced gradients. T1 rho shows promise as a contrast parameter in high-field MRH because it is capable of producing T2-like contrast without the susceptibility artifacts associated with T2-weighted images.

Pulsed magnetization transfer contrast for MR imaging with application to breast

The relative populations and transverse relaxation times of the solid-like hydrogen pool (PB and T2B) and the magnetization transfer (MT) rates between the solid-like and liquid-like hydrogen pools (kappa) have been determined for three different agar gel concentrations (2%, 4%, and 8% by weight) as well as excised fibroglandular breast tissue specimens. PB was determined to be .003(.001), .01(.002), .02(.01), and .06(.01); T2B was determined to be 13.0(.2), 14.0(.1), 14.5(.1) and 15.2(1.3) microseconds; and kappa was determined to be 0.78(.01), 1.15(.02), 2.00(.02), and 3.55(1.5) sec-1 for the 2%, 4%, and 8% agar gels and the fibroglandular tissue, respectively. The image signal intensities of a pulsed MTC-prepared gradient-echo imaging technique are predicted using these MT parameters and are shown to agree well with experimental data obtained from a clinical MR imaging system. This technique is shown to suppress signal intensity of fibroglandular breast tissue by 40%-50% without exceeding SAR limits (< or = 8 W/kg) and is helpful for visualizing lesions and silicone implants.

Spin lock magnetic resonance imaging in the differentiation of hepatic haemangiomas and metastases

Spin lock (SL) imaging technique, generating T1 rho-weighted images, was applied to the differentiation of hepatic haemangiomas from metastatic focal liver lesions. 17 haemangiomas and 16 metastases in 32 patients were imaged at the field-strength of 0.1 T using a multiple slice SL technique and a conventional gradient-echo (GRE) sequence with identical timing parametres. Spin lock effects of the hepatic lesions and different abdominal tissues were calculated. Images with adequate coverage of the liver and of good quality with few motion induced artefacts were acquired. A definite, statistically significant, difference was found between the SL-effects of hepatic haemangiomas and a liver metastases. Haemangiomas showed an SL effect of 46.6 +/- 3.4% and metastases of 56.2 +/- 5.8% (mean +/- SD, p < 0.0001). The multiple slice SL technique showed potential in distinguishing haemangiomas from metastatic liver lesions and should be considered as an alternative to the conventional T2 and magnetization transfer (MT) based methods.

Near-resonance spin-lock contrast

Spin-lock and spin-tip excitations are the two magnetization components created by the preparatory RF pulse of an MRI contrast enhancement sequence. Only spin-lock is inherently adiabatic, preserving spin alignment so that tissue-specific relaxation can generate desired saturation contrasts. Spin-tip is the rotating-frame oscillating excitation, and generally causes nonadiabatic loss of all detectable magnetization. Relative levels of spin-lock and spin-tip are important to understand as a function of the preparatory B1 delta amplitude, resonance frequency offset, delta, and the pulse waveform. These MR responses can be accurately analyzed theoretically and numerically by using Torrey's tipped coordinates to formulate Bloch's equations. At near-resonance offsets, (delta/gamma B1) less than 2.0, spin-lock contrast (SLC) depends strongly on T2, due to the nature of spin-lock T1 rho relaxation in the RF pulse interval. The relaxation rates 1/T1 rho and 1/T2 rho apply for active B1 delta, but remain linear combinations of ordinary (1/T1) and 1/T2) for motionally narrowed MR. The SLC increases rapidly as delta decreases below 2000 Hz; carefully chosen B1 delta rise times avoid spin-tip losses down to 150 Hz or less. The SL saturation enhances or multiplies any other indirect saturation effects that may be also present, such as magnetization transfer. A strong near-resonance SLC multiplier is advantageous for clinically practical MRI sequences that use short B1 delta pulses and fast SE multislice scan modes. Simulations based upon spin-lock/spin-tip theory and measured (T1,T2) yield excellent agreement with real MRI results for clinically practical fast multislice scans.

Off-resonance spin locking for MR imaging

Off-resonance spin locking is investigated as a low power method for achieving low field spin-lattice relaxation contrast using high field clinical MR imaging systems (e.g., 1.5 tesla). Spin-lattice relaxation times and equilibrium magnetizations in the off-resonance rotating frame (T1 rho(off), beta) were measured for tissue-mimicking phantom materials as a function of the ratio of the amplitude to the resonance offset of the spin-locking pulse (f1/delta). The phantom materials consisted of vegetable oil to simulate fat and two different gels containing 2% and 4% agar to simulate nonfatty tissues with different macromolecular compositions. These measurements were used to verify a signal strength equation for a multislice off-resonance spin-locking technique implemented on a clinical MR imaging system operating at 1.5 tesla. Although the oil showed little change in image contrast with increasing f1/delta, the two gels demonstrated a strong variation which provided improved discrimination compared to T1-weighted imaging. Off-resonance spin locking is suggested as a method for improving delineation of breast lesions and a preliminary clinical example from a patient volunteer is presented.

Analysis of discrete T2 components of NMR relaxation for aqueous solutions in hollow fiber capillaries

An analysis is presented of proton NMR T2 relaxation times measured for aqueous solutions in simple bundles of hollow fibers. The relaxation times are calculated with a two-compartment diffusive exchange model using the known relaxation times of the aqueous solutions and the fiber geometry. When the relaxation time outside the fibers is short (approximately 1 ms), three or more relaxation components are observed from this two compartment system, in agreement with the calculation. The amplitude and relaxation times of the third component are consistent with those of a diffusion-mediated mode, as suggested theoretically by Brownstein and Tarr (Phys. Rev. A 19, 2446 (1979)). The possible contribution of such modes to the multicomponent relaxation observed in tissues is discussed.

T1 rho dispersion imaging of diseased muscle tissue

T1 rho dispersion, or the frequency dependence of T1 relaxation in the rotating frame, was used for in vivo muscle tissue characterization in 13 patients with primary skeletal muscle disease and in eight normal subjects for comparison. T1 rho dispersion measurements represent a new approach to magnetic resonance tissue characterization, possibly reflecting the macromolecular constituents of tissue. A definite, statistically significant, difference was found between the relative T1 rho dispersion values of normal and diseased muscle tissue. T1 rho dispersion measurements and images may increase the accuracy of identification of diseased muscles. Early identification of affected muscles is important for accurate diagnosis by muscle biopsy.

On the application of ultra-fast RARE experiments

The ultra-fast application of the RARE experiment is described in detail, with special emphasis on its multifarious applications with preparation experiments that produce transverse magnetization. The factors affecting the temporal evolution of the magnetization during the experiment are described, and the implications for the slice profile when using a Gaussian refocusing pulse are experimentally examined. The choice of phase-encoding scheme for use with preparation experiments is discussed, as is the use of various phase-encoding schemes to reduce line broadening in the phase-encoding direction if a number of averages are acquired. An explanation for the decomposition of the echo are into two components if the read gradient is imbalanced is given, and the experimental conditions necessary for the coherent addition of these two echo groups are described. An alternative sequence that removes one of these groups from the acquisition window is proposed. The sensitivity of the sequence to flow and motion is investigated, and the drastic loss of signal in this situation explained. The in vivo and in vitro application of preparation experiments leading to the accurate measurement of T1, T2, diffusion constant, and magnetization transfer characteristics is presented. The implementation of zoom-imaging using spin- and stimulated-echo preparation is described, and 3D in vivo spin-echo zoom images are presented. Simple phantom experiments demonstrating the feasibility of chemical-shift selective and spectroscopic imaging are also given.

Breast coil design for low-field MRI

Four double-breast coils were designed for the low-field resistive magnet MR imaging of female breasts at 0.02, 0.04, and 0.1 T. The signal-to-noise ratio is optimized by shaping the coils, by constructing two different size coils at 0.02 T, and by using special wire for the turns of the solenoidal coils. The approximate linear dependence of the SNR on the Larmor frequency is estimated.

Magnetization transfer contrast in fat-suppressed steady-state three-dimensional MR images

We demonstrate that magnetization transfer contrast can be used to improve the diagnostic utility of fat-suppressed steady-state three-dimensional gradient-recalled images. Fat suppression is achieved using a "jump-return" pair of contiguous shaped pulses. No time interval exists between the pulses, and no RF echo is generated. The sequence normally produces images with "density" weighting. Preparation of the spin magnetization with off-resonance frequency-selective excitation creates magnetization transfer contrast which attenuates signal intensity in proportion to the exchange rate of magnetization from free water with magnetization from water bound to macromolecules or protons that have restricted mobility. The resulting images have excellent fat suppression with low sensitivity to motion since no subtraction is used. In addition, the mechanism of signal attenuation is independent of paramagnetic effects, and addition of Gd-DTPA produces signal enhancement from vascularized regions of tissue. Examples are presented for the knee and breast, where the observation of pathology with signal enhancement from Gd-DTPA is improved over conventional 3D fat-suppressed images.

T1 rho dispersion imaging and localized T1 rho dispersion relaxometry: application in vivo to mouse adenocarcinoma

The dispersion (frequency dependence) of the spin-lattice relaxation time in the rotating frame, T1 rho, is considered for tissue characterization. Methods for the volume-selective determination of the proper T1 rho dispersion and for imaging of parameters characterizing this frequency dependence are described. On- and off-resonance versions of the techniques are demonstrated. In vitro studies of excised rat tissues and in vivo applications to mice with implanted adenocarcinoma are reported. T1 rho dispersion images show clear contrasts of the malignant tissue, whereas muscle tissue is completely suppressed. No contrast agent is required. The measuring time is only twice as long as that for conventional magnetic resonance images. The results suggest that the T1 rho dispersion is less susceptible to the biological variability than the absolute values of the relaxation times.