Spiral T1 Spin-Echo for Routine Postcontrast Brain MRI Exams: A Multicenter Multireader Clinical Evaluation.

BACKGROUND AND PURPOSE: Spiral MR imaging has several advantages compared with Cartesian MR imaging that can be leveraged for added clinical value. A multicenter multireader study was designed to compare spiral with standard-of-care Cartesian postcontrast structural brain MR imaging on the basis of relative performance in 10 metrics of image quality, artifact prevalence, and diagnostic benefit. MATERIALS AND METHODS: Seven clinical sites acquired 88 total subjects. For each subject, sites acquired 2 postcontrast MR imaging scans: a spiral 2D T1 spin-echo, and 1 of 4 routine Cartesian 2D T1 spin-echo/TSE scans (fully sampled spin-echo at 3T, 1.5T, partial Fourier, TSE). The spiral acquisition matched the Cartesian scan for scan time, geometry, and contrast. Nine neuroradiologists independently reviewed each subject, with the matching pair of spiral and Cartesian scans compared side-by-side, and scored on 10 image-quality metrics (5-point Likert scale) focused on intracranial assessment. The Wilcoxon signed rank test evaluated relative performance of spiral versus Cartesian, while the Kruskal-Wallis test assessed interprotocol differences. RESULTS: Spiral was superior to Cartesian in 7 of 10 metrics (flow artifact mitigation, SNR, GM/WM contrast, image sharpness, lesion conspicuity, preference for diagnosing abnormal enhancement, and overall intracranial image quality), comparable in 1 of 10 metrics (motion artifacts), and inferior in 2 of 10 metrics (susceptibility artifacts, overall extracranial image quality) related to magnetic susceptibility (P < .05). Interprotocol comparison confirmed relatively higher SNR and GM/WM contrast for partial Fourier and TSE protocol groups, respectively (P < .05). CONCLUSIONS: Spiral 2D T1 spin-echo for routine structural brain MR imaging is feasible in the clinic with conventional scanners and was preferred by neuroradiologists for overall postcontrast intracranial evaluation.

A Spiral Spin-Echo MR Imaging Technique for Improved Flow Artifact Suppression in T1-Weighted Postcontrast Brain Imaging: A Comparison with Cartesian Turbo Spin-Echo.

BACKGROUND AND PURPOSE: A challenge with the T1-weighted postcontrast Cartesian spin-echo and turbo spin-echo brain MR imaging is the presence of flow artifacts. Our aim was to develop a rapid 2D spiral spin-echo sequence for T1-weighted MR imaging with minimal flow artifacts and to compare it with a conventional Cartesian 2D turbo spin-echo sequence. MATERIALS AND METHODS: T1-weighted brain imaging was performed in 24 pediatric patients. After the administration of intravenous gadolinium contrast agent, a reference Cartesian TSE sequence with a scanning time of 2 minutes 30 seconds was performed, followed by the proposed spiral spin-echo sequence with a scanning time of 1 minutes 18 seconds, with similar spatial resolution and volumetric coverage. The results were reviewed independently and blindly by 3 neuroradiologists. Scores from a 3-point scale were assigned in 3 categories: flow artifact reduction, subjective preference, and lesion conspicuity, if any. The Wilcoxon signed rank test was performed to evaluate the reviewer scores. The t test was used to evaluate the SNR. The Fleiss κ coefficient was calculated to examine interreader agreement. RESULTS: In 23 cases, spiral spin-echo was scored over Cartesian TSE in flow artifact reduction (P < .001). In 21 cases, spiral spin-echo was rated superior in subjective preference (P < .001). Ten patients were identified with lesions, and no statistically significant difference in lesion conspicuity was observed between the 2 sequences. There was no statistically significant difference in SNR between the 2 techniques. The Fleiss κ coefficient was 0.79 (95% confidence interval, 0.65-0.93). CONCLUSIONS: The proposed spiral spin-echo pulse sequence provides postcontrast images with minimal flow artifacts at a faster scanning time than its Cartesian TSE counterpart.

Diffusion-weighted imaging provides support for secondary neuronal damage from intraparenchymal hematoma.

It is controversial whether an intracerebral hematoma (ICH) causes ischemia of surrounding brain. By virtue of its high sensitivity to acute cerebral infarction, diffusion-weighted imaging (DWI) helps answer this question. We used this technique to assess the parenchyma surrounding ICH for restricted diffusion. Echoplanar DWI (b 1000 s/mm(2)) and conventional MRI sequences were performed in 30 subjects (symptom duration 7-75 h) with primary ICH, mean volume: 13+/-15 cm(3). We calculated mean apparent diffusion coefficients (ADC) within high signal regions around the hematoma on DWI or T2-weighted images and within the ICH itself, comparing them to the contralateral brain. We used the Student's t -test to examine for differences between these regions and linear regression to relate changes to the age of the ICH. A thin rim of high signal on DWI and a wider rim on T2-weighted images surrounded all hematomas. The ADC within the rim on DWI showed a maximum reduction of 40%, in two patients imaged within 10 h of symptom onset. They rose during the first day (r(2)=0.84; P <0.03) and then showed a mild decrease, becoming the same as ADC in other areas of the brain (r(2)=0.5; P <0.03). The rim on T2-weighting showed a mean increase of 50% and ADC within the ICH were reduced by a mean of 38%; these variations showed no relationship with ICH age and no group showed any relationship with ICH size. The ADC within the three regions was significantly different from each other. The presence of restricted diffusion in the parenchyma surrounding ICH provides support for secondary neuronal damage.

PROPELLER MRI: clinical testing of a novel technique for quantification and compensation of head motion.

While head motion is considered a significant problem in magnetic resonance imaging (MRI), there is no data to quantify its extent, severity, or effect on image quality. PROPELLER (Periodically Rotated Overlapping ParallEL Lines with Enhanced Reconstruction) MRI offers a novel means of quantifying and compensating for head motion. We performed axial T2-weighted PROPELLER (motion corrected: P-CR; uncorrected: P-UNCR) and conventional MRI (CONV), with equal scan times, in five normal volunteers and 35 clinical subjects. Volunteers were examined lying still and performing two separate head movements (shake "no" and nod "yes") to assess detection and compensation of in-plane motion by PROPELLER MRI. Images were examined by three radiologists for motion artifact and for overall image quality. Head rotation and translation was detected in all subjects during each slice acquisition, with expected changes occurring with volunteer head motion. Motion artifact was less commonly seen on PROPELLER than CONV MR (chi(2) test P < 0.001). PROPELLER was preferred over CONV in all subjects (P < 0.05) and P-CR was judged superior to P-UNCR (P = 0.02). Intracranial pathology was equally or better demonstrated with PROPELLER. PROPELLER MRI offers a means of quantifying head motion, reducing motion artifact, and improving image quality.

Limits of time-of-flight magnetic resonance angiography.

Many of the limits to current time-of-flight (TOF) magnetic resonance (MR) angiographic methods, including flow voids from turbulent flow, flow artifacts from pulsatile flow, and oblique flow displacement, are not inherent to the TOF enhancement mechanism and will be eliminated with technological advancements. Slowly flowing blood and physiological motion will remain an obstacle to these methods, although technological advancements will improve on current methods. Future TOF MR angiographic methods will prove to be a cost-effective competitor to MR angiographic methods requiring contrast agents in many applications.

Evidence for cytotoxic edema in the pathogenesis of cerebral venous infarction.

BACKGROUND AND PURPOSE: The pathogenesis of cerebral venous infarction (CVI) remains controversial, with uncertainty over whether cytotoxic edema plays a role. Recent animal studies have shown that cytotoxic edema reliably occurs in acute CVI and precedes the onset of vasogenic edema. Our hypothesis was that cytotoxic edema would also occur in acute human CVI and would be detectable as an area of restricted diffusion on diffusion-weighted images. METHODS: Twelve subjects with acute cerebral venous thrombosis confirmed by MR venography underwent both conventional MR and echo-planar diffusion-weighted imaging (maximum diffusion sensitivity [b=1000 s/mm(2)]). Images were examined for areas of CVI that were identified as T2 hyperintensity, diffusion hyperintensity, or hemorrhage. The percent change in apparent diffusion coefficient (ADC) and T2 signal as well as the T2/diffusion volume were calculated within areas of edematous CVI. Regression techniques were used to examine the relationship of these variables to symptom duration. RESULTS: Ten regions of CVI were detected in seven subjects, all showing T2 hyperintensity. Two of these regions were predominantly hemorrhagic and did not display diffusion hyperintensity. The remaining eight regions displayed diffusion hyperintensity that was associated with a decreased ADC. ADC values increased with symptom duration (r(2) = 0.96; P <.006). Both T2 hyperintensity and T2/diffusion volume peaked approximately 2 days after symptom onset. CONCLUSION: Restricted water diffusion suggesting cytotoxic edema is commonly found in subjects with acute CVI and decreases over time. This supports an important etiologic role for cytotoxic edema in the pathogenesis of CVI.

Pulse waveform analysis of arterial compliance: relation to other techniques, age, and metabolic variables.

To assess the physiologic and clinical relevance of newer noninvasive measures of vascular compliance, computerized arterial pulse waveform analysis (CAPWA) of the radial pulse was used to calculate two components of compliance, C1 (capacitive) and C2 (oscillatory or reflective), in 87 normotensive (N1BP, n = 20), untreated hypertensive (HiBP, n = 21), and treated hypertensive (HiBP-Rx, n = 46) subjects. These values were compared with two other indices of compliance, the ratio of stroke volume to pulse pressure (SV/PP) and magnetic resonance imaging (MRI)-based aortic distensibility; and were also correlated with demographic and biochemical values. The HiBP subjects displayed lower C1 (1.34 +/- 0.09 v. 1.70 +/- 0.11 mL/mm Hg, significance [sig] = .05) and C2 (0.031 +/- 0.003 v 0.073 +/- 0.02 mL/mm Hg, sig = .005) than N1BP subjects. This was not true for C1 (1.64 +/- 0.08 mL/mm Hg) and C2 (0.052 +/- 0.005 mL/mm Hg) values in HiBP-Rx subjects. The C1 (r = 0.917, P < .0001) and C2 (r = 0.677, P < .0001) were both closely related to SV/PP, whereas C1 (r = 0.748, P = .002), but not C2, was significantly related to MRI-determined aortic distensibility. Among other factors measured, age exerted a strong negative influence on both C1 (r = -0.696, P < .0001) and C2 (r = -0.611, P < .0001) compliance components. Positive correlations were observed between C1 (r = 0.863, P = .006), aortic distensibility (r = 0.597, P = .19) and 24-h urinary sodium excretion, and between C1- and MR spectroscopy-determined in situ skeletal muscle intracellular free magnesium (r = 0.827, P = .006), whereas C2 was inversely related to MRI-determined abdominal visceral fat area (r = -0.512, P = .042) and fasting blood glucose (r = -0.846, P = .001). Altogether, the close correspondence between CAPWA, other compliance techniques, and known cardiovascular risk factors suggests the clinical relevance of CAPWA in the assessment of altered vascular function in hypertension.

Rapid method for deblurring spiral MR images.

A method for fast off-resonance frequency deblurring of spiral MR images is presented. The method utilizes image-space deconvolution. The off-resonance phase is approximated as a separable quadratic function to allow rapid one-dimensional deconvolution with a small compromise in accuracy. The method is used to deblur an MR angiographic image to illustrate its effectiveness.

Neonatal hypoxic-ischemic encephalopathy: detection with diffusion-weighted MR imaging.

BACKGROUND AND PURPOSE: Although diffusion-weighted imaging has been shown to be highly sensitive in detecting acute cerebral infarction in adults, its use in detecting neonatal hypoxic-ischemic encephalopathy (HIE) has not been fully assessed. We examined the ability of this technique to detect cerebral changes of acute neonatal HIE in different brain locations. METHODS: Fifteen MR examinations were performed in 14 neonates with HIE (median age, 6.5 days; range, 2-11 days). Imaging comprised conventional T1-weighted, proton density-weighted, and T2-weighted sequences and echo-planar diffusion-weighted sequences. The location, extent, and image timing of ischemic damage on conventional and diffusion-weighted sequences and apparent diffusion coefficient (ADC) maps were compared. RESULTS: Although conventional sequences showed cerebral changes consistent with ischemia on all examinations, diffusion-weighted imaging showed signal hyperintensity associated with decreased ADC values in only seven subjects (47%). All subjects with isolated cortical infarction on conventional sequences had corresponding hyperintensity on diffusion-weighted images and decreased ADC values, as compared with 14% of subjects with deep gray matter/perirolandic cortical damage. The timing of imaging did not significantly alter diffusion-weighted imaging findings. CONCLUSION: Diffusion-weighted imaging, performed with the technical parameters in this study, may have a lower correlation with clinical evidence of HIE than does conventional MR imaging. The sensitivity of diffusion-weighted imaging in detecting neonatal HIE appears to be affected by the pattern of ischemic damage, with a lower sensitivity if the deep gray matter is affected as compared with isolated cerebral cortex involvement.

Reconstructing MR images from undersampled data: data-weighting considerations.

Data which are sampled more densely than the Nyquist limit in k-space are weighted prior to reconstruction by the inverse of the local sampling density. This work considers the effects of weighting data that are sampled less densely than the Nyquist limit. It specifically analyzes azimuthally undersampled projection reconstruction, variable density spirals, and variable density phase encoding. Effects on resolution, aliasing, and SNR are given. Higher resolution is obtained by weighting undersampled data according to the inverse of sampling density, while better SNR and less aliasing artifact are obtained by weighting undersampled data uniformly. Magn Reson Med 43:867-875, 2000.

Motion correction with PROPELLER MRI: application to head motion and free-breathing cardiac imaging.

A method for motion correction, involving both data collection and reconstruction, is presented. The PROPELLER MRI method collects data in concentric rectangular strips rotated about the k-space origin. The central region of k-space is sampled for every strip, which (a) allows one to correct spatial inconsistencies in position, rotation, and phase between strips, (b) allows one to reject data based on a correlation measure indicating through-plane motion, and (c) further decreases motion artifacts through an averaging effect for low spatial frequencies. Results are shown in which PROPELLER MRI is used to correct for bulk motion in head images and respiratory motion in nongated cardiac images. Magn Reson Med 42:963-969, 1999.

An optimized center-out k-space trajectory for multishot MRI: comparison with spiral and projection reconstruction.

A method for collecting MRI data on a set of rotated trajectories that begin at the center of k-space is outlined. It is theoretically slightly faster, slightly less susceptible to off-resonance and motion-induced phase, and produces images with slightly better signal-to-noise ratio than methods using Archimedean spiral trajectories, particularly for a short sampling duration. It also produces more accurate images than those of projection reconstruction methods, which are significantly undersampled azimuthally. This method may be most useful when imaging areas with large inhomogeneities (e.g., near metallic implants), short T(2) species, or high turbulence (e.g., gas imaging). Magn Reson Med 42:714-720, 1999.

Flow effects in localized quadratic, partial Fourier MRA.

A pulse sequence for inflow-enhanced magnetic resonance angiography, including localized quadratic encoding, partial-Fourier slice selection, and spiral in-plane encoding, is analyzed. The through-plane encoding method is discussed in a space-spatial frequency context to illustrate some of its properties. This pulse sequence has the advantages of being faster and more robust to turbulent flow than conventional inflow-enhanced methods. Simulations show the effect of different parameters on the modulation-transfer function of the resulting images. A flow phantom is used to verify some of the simulation results.

Effects of interleaf order for spiral MRI of dynamic processes.

The effects of the temporal order in which spiral interleaves are collected are discussed, in the context of artifacts from moving or changing objects. Simulations and in vivo experiments demonstrate the properties of four different ordering methods. Specific applications discussed include cardiac and interventional magnetic resonance imaging, as well as inflow and contrast-enhanced magnetic resonance angiography.

Sampling density compensation in MRI: rationale and an iterative numerical solution.

Data collection of MRI which is sampled nonuniformly in k-space is often interpolated onto a Cartesian grid for fast reconstruction. The collected data must be properly weighted before interpolation, for accurate reconstruction. We propose a criterion for choosing the weighting function necessary to compensate for nonuniform sampling density. A numerical iterative method to find a weighting function that meets that criterion is also given. This method uses only the coordinates of the sampled data; unlike previous methods, it does not require knowledge of the trajectories and can easily handle trajectories that "cross" in k-space. Moreover, the method can handle sampling patterns that are undersampled in some regions of k-space and does not require a post-gridding density correction. Weighting functions for various data collection strategies are shown. Synthesized and collected in vivo data also illustrate aspects of this method.

Basic spin physics.

Magnetic resonance imaging is fundamentally a measurement of the magnetism inherent in some nuclear isotopes; of these the proton, or hydrogen atom, is of particular interest for clinical applications. The magnetism in each nucleus is often referred to as spin. A strong, static magnetic field B0 is used to align spins, forming a magnetic density within the patient. A second, rotating magnetic field B1 (RF pulse) is applied for a short duration, which rotates the spins away from B0 in a process called excitation. After the spins are rotated away from B0, the RF pulse is turned off, and the spins precess about B0. As long as the spins are all pointing in the same direction at any one time (have phase coherence), they act in concert to create rapidly oscillating magnetic fields. These fields in turn create a current in an appropriately placed receiver coil, in a manner similar to that of an electrical generator. The precessing magnetization decays rapidly in a duration roughly given by the T2 time constant. At the same time, but at a slower rate, magnetization forms again along the direction of B0; the duration of this process is roughly expressed by the T1 time constant. The precessional frequency of each spin is proportional to the magnetic field experienced at the nucleus. Small variations in this magnetic field can have dramatic effects on the MR image, caused in part by loss of phase coherence. These magnetic field variations can arise because of magnet design, the magnetic properties (susceptibility) of tissues and other materials, and the nuclear environment unique to various sites within any given molecule. The loss of phase coherence can be effectively eliminated by the use of RF refocusing pulses. Conventional MR imaging experiments can be characterized as either gradient echo or spin echo, the latter indicating the use of a RF refocusing pulse, and by the parameters TR, TE, and flip angle alpha. Tissues, in turn, are characterized by their individual spin density, M0, and by the T1, T2, and T2* time constants. Knowledge of these parameters allows one to calculate the resulting signal from a given tissue for a given MR imaging experiment.

Asymmetric sampling along k(slice-select) in two-dimensional multislice MRI.

Reduction of the slice-select refocusing gradient in two-dimensional multislice imaging results in asymmetry of the k-space representation of collected data along the slice-select direction. Standard methods of partial Fourier reconstruction developed for other methods of asymmetric k-space sampling can be used to reconstruct these data with final through-plane resolution smaller than the collected slice thickness. This method can be used for reducing scan time in the same manner as asymmetric sampling in the phase-encoded direction. In addition, the reduced refocusing gradient reduces minimum TE and motion artifacts in the same manner as for asymmetric sampling in the frequency-encoded direction (fractional echoes). Results using a resolution phantom and a flow phantom illustrate these concepts.

Direct magnetic resonance determination of aortic distensibility in essential hypertension: relation to age, abdominal visceral fat, and in situ intracellular free magnesium.

To investigate the contribution of vascular compliance to essential hypertension (EH), we developed magnetic resonance imaging (MRI) techniques to directly measure aortic distensibility (AD) in the ascending and descending thoracic and abdominal aorta of fasting normal (n= 10) and EH (n=20) subjects. These results were compared with concurrent MR-based measurements of left ventricular mass index (LVMI) and abdominal subcutaneous and visceral fat and with 31P-MR spectroscopic measurement of in situ intracellular free magnesium levels (Mgi) in brain and skeletal muscle. Aortic distensibility in EH was consistently and significantly reduced at all measured sites (2.5+/-0.4, 2.2+/-0.4, 2.3+/-0.4 versus 7.0+/-1.6, 5.1+/-0.3, 7.3+/-0.8 mm Hg(-1) x 10(-3), P<.05), as was Mgi in the brain (284+/-22 versus 383+/-34 micromol/L, P<.05) and skeletal muscle (397+/-10 versus 527+/-36 micromol/L, P<.05). For all subjects, systolic blood pressure (r=-.662, P<.0001) and LVMI (r=-.484, P<.01) were inversely related to AD. AD and brain Mgi were inversely related to age (AD, r=-.792, P<.0001; brain Mgi: r=-.673, P<.05). AD was inversely related to fasting blood glucose (r=-.413, P<.05) and to abdominal visceral fat (r=-.416, P<.05) but not to body mass index (BMI: r=-.328, P=NS) or subcutaneous fat (r=-.157, P=NS). AD was also significantly and positively related to in situ Mgi, both in the brain and skeletal muscle (brain: r=.712, P<.01; skeletal muscle: r=.632, P<.01). We conclude that (1) MR techniques can be used to coordinately and noninvasively assess cardiac, vascular, metabolic, and ionic aspects of hypertensive disease in humans; (2) increased systolic blood pressure and LVMI in EH may at least in part result from decreased AD; (3) decreased Mgi contributes to arterial stiffness in hypertension and may help to explain the characteristic age-related decreases in AD; and (4) decreased AD may be one mechanism by which abdominal visceral fat contributes to cardiovascular risk.

Analysis of localized quadratic encoding and reconstruction.

Localized quadratic (LQ) encoding is a method of collecting multislice MRI data sets with greatly overlapping excited slices in such a way that the slice overlap can be deconvolved. This results in reconstructed data with resolution equal to the center-to-center slice spacing, regardless of the amount of excited slice overlap. LQ encoding is analyzed using the modulation transfer function of the encoding and reconstruction process. This allows analysis of many aspects of the technique in a well established theoretical framework. Many different characteristics of the method are explored, including excited slice profiles, required RF magnitudes and specific absorption rate, signal to noise ratio, signal dynamic range, reconstruction artifacts, sensitivity to motion, saturation and inflow effects, modulation transfer function shifting, and off-resonance artifacts. It is suggested that this technique is best suited for applications currently using multiple thin-slab three-dimensional encoding.

Early detection of treatment response by diffusion-weighted 1H-NMR spectroscopy in a murine tumour in vivo.

Nuclear magnetic resonance (NMR) non-invasively measures the apparent diffusion coefficient (ADC) of water, which is sensitive to the biophysical characteristics of tissue. Because anti-cancer treatment alters tumour pathophysiology, tumour ADC may be altered by treatment. In order to test this hypothesis, ADC was measured in s.c. implanted murine RIF-1 tumours before and up to 9 days after treatment with cyclophosphamide. A dose-dependent, reversible increase in tumour ADC was observed after cyclophosphamide treatment, which is consistent with an increase in the fraction of interstitial water due to treatment-induced cell death. Because tumour water ADC is increased substantially at a time when there is no change in tumour volume for a dose which produces minimal cell kill, its measurement could provide a novel means for early detection of response to anti-cancer therapy. If the changes in ADC observed in the present study are evident for commonly used anti-cancer therapies in different tumour types and specific to a therapeutic response, the approach could be broadly applicable as a response predictor since magnetic resonance imaging can be used to measure ADC in human tumours.