Biomedical subjects
Publications and source records attributed to F W Wehrli.
In trabecular bone the NMR linewidth is found to be governed by the magnetic field inhomogeneity arising from the difference in magnetic permeability between bone and marrow. T2* was measured in vivo in the distal femur by interferometry and localized spectroscopy. Both methods showed T2* to be inversely related to trabecular plate density.
Temporally resolved velocity measurements in the popliteal arteries of 11 healthy subjects were obtained by means of magnetic resonance (MR) imaging with use of the Fourier flow-encoding technique. Excellent agreement with corresponding Doppler ultrasonography (US) data (r = .97, slope = 0.99, intercept = -1.5 cm/sec) was demonstrated over the entire velocity range from 50 to -20 cm/sec. The method was rapid and its implementation straightforward. Further, MR imaging was shown to provide the intraluminal velocity distribution relevant for the determination of true flow rates, not obtainable with Doppler US.
A new approach to probe the structure of trabecular bone in the vertebral bodies in humans was evaluated, and preliminary data are presented. The proposed method is based on the hypothesis that the presence of two physical phases--bone and bone marrow--causes a magnetic field distribution across the imaging voxel. The resulting spread in resonance frequency produces line broadening, which is measured as the decay rate of the region of interest signal intensity that has the properties of an interferogram. The interferogram is the result of two principal chemically shifted components of bone marrow--fat and water--getting in and out of phase with one another while being attenuated by T2* processes from the magnetic field distribution within the measuring volume. The time constant for the decay (T2*) can then be obtained by means of curve-fitting techniques. T2* in healthy persons is found to increase slightly with age. However, patients with osteoporosis (low bone mineral density and/or spine compression fractures) have significantly prolonged T2* values, which are interpreted as arising from an increase in the intertrabecular space.
Early results are reported of hepatic neoplasms studied with dynamic gadolinium-diethylenetriaminepentaacetate (Gd-DTPA)-enhanced magnetic resonance imaging (MRI). The purpose of this study was to evaluate the potential of the Gd enhancement pattern for characterizing these neoplasms and to compare the performance of spin-echo (SE) to SE plus dynamic Gd-enhanced gradient-echo (GRE) pulse sequences. Forty-two patients with hepatic neoplasms were examined at 1.5 T field strength. In each patient, short and long repetition time/echo time (TR/TE) SE images were obtained, followed by pre- and post-Gd-DTPA (0.1 mmol/kg body wt), single-slice, breath-hold (13 s/scan) GRE images, which were serially acquired less than or equal to 12 min postinjection. The patterns of contrast enhancement of the various hepatic lesions were documented and analyzed. The time to peak Gd signal enhancement-to-noise ratio (SE/N), contrast-to-noise ratio (C/N), contrast (defined as the signal intensity ratio [SIR]), as well as the peak values of these quantities, were determined. The C/N and SIR values on the short and long TR/TE SE and pre-Gd GRE images for all hepatic neoplasms were also obtained. The discriminating abilities (hemangiomas vs. malignant neoplasms) of these quantities were analyzed statistically. In addition, the impact of lesion characterization on the SE versus the SE plus the Gd-enhanced GRE scans was assessed by means of a blinded reader study. Malignant hepatic neoplasms could be differentiated from hemangiomas by visual inspection of their enhancement patterns (p = 0.0009), by the time to peak C/N on Gd-enhanced images (p = 0.0002), and by the magnitude of such peak (p = 0.02). Combined SE + Gd-enhanced GRE scans afforded minor, though statistically significant (p less than 0.01), improvement of the accuracy in differentiating benign from malignant hepatic neoplasms. Late scans (12 min post-Gd) may be particularly useful in identifying hemangiomas that, unlike other neoplasms, have a significant high signal of their enhancing portions on such images.
This article provides a brief overview of one of the two fundamental physical principles which lead to a modulation of the magnetic resonance signal of moving spins: the time-of-flight effect. It then discusses some of its characteristic manifestations in spin-echo imaging and reviews strategies for exploiting the effect to quantitate vascular flow.
Fifty-four independent scans were performed in two volunteers covering one anatomic region in each (the brain and knee) with the purpose of ascertaining the agreement between predicted and measured signal-to-noise ratios (SNR). Systematically varied parameters were number of excitations (NEX), field of view (FOV), section thickness (dz), and the number of phase-encoding steps (Ny). Correlation coefficients of measured versus predicted SNR were 0.82 and 0.86, respectively, in the anatomies studied. Significantly improved correlations were found for data subpopulations in which NEX was held constant. To assess the criteria guiding reader preference, a blinded study was performed in which radiologists were asked to rate images from least to most desirable. In order to quantitatively determine the criteria for reader preference, plots of mean rating versus SNR, voxel volume, and an image quality index [IQI = SNR/(voxel volume)] were performed. The latter was found to be a better predictor of reader preference than either SNR or spatial resolution alone. The data suggests T1-weighted scan protocols yielding SNR of approximately 20 are preferable with any excess SNR being traded for smaller voxel size or shorter scan times.
Routine evaluation of axial MR images of the cervical spine with high-intensity CSF (long TR/TE spin-echo or gradient-echo images) revealed apparent narrowing of the cord's anteroposterior diameter when these images were compared with corresponding postmyelography CT scans. This discrepancy was believed to be due to the truncation artifact at the CSF-cord boundary. To examine the truncation effect, we compared cord diameters in 12 patients on postmyelography CT scans and MR images and then compared these with MR scans of normal volunteers and of an agar-saline spine phantom. There was an artifactual diminution of the cord diameter in the 128-step phase-encoding axis of the 128 x 256-matrix MR scan as compared with the diameter of the cord in the patients' postiohexol CT scans and in the 256 phase-encoded axis MR scan in the volunteer study. A similar discrepancy was noted in the spine phantom study, in which the cord diameter in the 256-step phase-encoded MR scan, the CT scan, and direct measurement exceeded that in the 128-step phase-encoded axis MR scan. The range of differences between the measurements was as large as 2.3 mm (patients), 1.7 mm (volunteers), and 1.8 mm (phantom) for the three studies. In all three studies, varying the photographic window width and level produced variation in the apparent cord diameter of up to 1.5 mm. To eliminate this effect, the cord diameters in the phantom and the normal control subjects were measured at identical window levels. The truncation artifact, coupled with standard window settings used in photography, may lead to inaccurate display of the diameter of the cervical spinal cord.
Spatial misregistration of signal recovered from flowing spins within vascular structures is a common phenomenon seen in MR imaging of the CNS. The condition is displayed as a bright line or dot offset from the true anatomic location of the lumen of the imaged vessel. Its origin is the time delay between application of the phase- and frequency-encoding gradients used to locate spins within the plane of section. The principal condition necessary for the production of spatial misregistration is flow oblique to the axis of the phase-encoding gradient. Flow-related enhancement (entry slice phenomenon), even-echo rephasing, and gradient-moment nulling contribute to the production of the bright signal of spatial misregistration. Familiarity with the typical appearance of flow-dependent spatial misregistration permits confirmation of a vessel's patency; identification of the direction of flow; estimation of the velocity of flow; and differentiation of this flow artifact from atheromas, dissection, intraluminal clot, and artifacts such as chemical shift.
Explore the source record for details and available documents.
Measurements of proton T1 and T2 were performed on GdCl3 solutions (20 less than T2 less than 500 msec, 90 less than T1 less than 1000 msec) on large-bore NMR imaging systems operating at 1.0T and 1.5T. CPMG multi-echo (ME), multiple saturation recovery (MSR) and modified fast inversion recovery (MFIR) pulse sequences as well as a sequence that combines and interleaves T1 and T2 weighted data acquisition (which we call "multiple saturation-recovery multiple-echo" (MSRME) were used. The relaxation data are compared to those obtained on a small bore NMR spectrometer operated at 1.5T. T1 and T2 values for the solutions were found to be the same within 10% for the two fields. Reproducibility of measurements of T1, T2 and the unnormalized spin density of the solutions was better than 5%. Systematic errors, amenable to correction through calibration, are noted in the imager T1 and T2 values. T1 and T2 values for some typical neural tissues at 1.5T and body tissue at 1.0T for human volunteers were obtained and are tabulated.
Explore the source record for details and available documents.
The authors present a method for obtaining magnetic resonance (MR) images of intra- and extracranial vessels from thin contiguous transaxial sections. A section-selective gradient refocusing pulse sequence with a short repetition time caused flow-related enhancement from spins that flowed perpendicular to the transaxial sections. The signal was further enhanced by means of flow compensation gradients to rephase any phase shifts resulting from moving spins in the presence of the imaging gradients. Coronal and sagittal sections, reformatted from multiple transaxial sections, are shown to have excellent vessel contrast without the use of contrast material. These images were obtained in 12 minutes of acquisition time from as many as 60 sections of 3-mm thickness. Such a technique shows significant promise for MR angiography.
This article addresses the dependence of the image characteristics in MR imaging of the orbit. Among the variables discussed are signal-to-noise (SNR), field of view (FOV), spatial resolution, and their interdependence. It is further shown how image contrast can be controlled by judicious choice of pulse timing parameters and how contrast depends on the intrinsic tissue MR parameters T1, T2, and proton density.
The reproducibility of T1, T2, and proton density, measured in phantoms and the human brain was evaluated by proton imaging techniques. The sequence used to derive T1 and density values was a multiple-saturation recovery which consists of four pairs of 90 degrees pulses, followed by a 180 degrees phase reversal pulse, generating four T1-weighted images. T2 was derived from a multiple-echo sequence, generating four T2-weighted images. The data were analyzed by fitting the pixel intensities to the respective equations by means of nonlinear multiparameter least-squares analysis. Short-term reproducibility between four consecutive scans was evaluated to be 1-4% depending on location with a phantom covering the entire span of physiologic T1 and T2 values. A second phantom containing a series of identical samples served to study the dependence of the apparent T1 and T2 on position, both radially and axially, with respect to magnet isocenter. Reproducibility across the field of view was found to be better than 7% (T1 and T2). This phantom was further used to evaluate effects of long-term reproducibility, which at each location varied from 5-14% (T1) and 2-10% (T2). Finally, interinstrument reproducibility, tested by means of the same protocol on three different instruments, all operating at the same magnetic field and using largely identical hardware for each location, was found to be 1-14% (T1) and 2-10% (T2). The positional dependence of the apparent relaxation times appears to be systematic and may be due to variations in the effective field, caused by magnet and rf inhomogeneity. Finally, brain tissue relaxation and spin-density data were determined using the same protocol in 37 scans performed on 27 normal volunteers. The tissues analyzed were putamen, thalamus, caudate nucleus, centrum semiovale, internal capsule, and corpus callosum. Excellent accordance was further obtained between left and right hemispheres.
The effects of section separation on image contrast and calculated T1 relaxation times were investigated in healthy volunteers and a phantom using an early commercial version magnetic resonance imaging system. The effects are explained qualitatively on the basis of side lobes of excitation occurring outside the selected section resulting in reduction of the time permitted for T1 relaxation. The options for dealing with imperfect section selection, including separation of the sections (i.e., leaving gaps) and nonsequential excitation, are illustrated and the trade-offs involved in each explained.
A novel magnetic resonance flow-imaging technique is presented and its suitability evaluated for both qualitative and quantitative imaging of flow. The method is derived from a selective saturation-recovery scheme consisting of a tagging and detection pulse followed by a bipolar read gradient. The detrimental phase effects causing signal loss at fast flow are shown to be greatly reduced because of the absence of a 180 degrees pulse and its associated section-selection gradient. The second loss mechanism intrinsic to 180 degrees spin echoes, the washout of excited spins between excitation and detection pulse, likewise is not present with the discussed technique. Assuming a parabolic flow profile, the authors calculated the signal evolution curve and found it to be in agreement with the experimental washout curve. The technique is shown to provide high-intensity signals for arteries such as carotid and vertebral arteries. Arteries and veins can be differentiated by judiciously choosing interpulse intervals or by alternating selective and nonselective tagging pulses.
Twenty-one patients with intraocular disease were studied by magnetic resonance (MR) imaging and computed tomography (CT). In 13 cases, malignant uveal melanoma was considered the likely diagnosis. Both imaging methods were accurate in determining the location and size of uveal melanomas. MR imaging was superior for the assessment of possible associated retinal detachment, for assessment of vitreous change, and for differentiating uveal melanoma from choroidal hemangioma and choroidal detachment. A case of retinal gliosis could not be differentiated from uveal melanoma by either technique. Uveal melanomas appeared as hyperintense lesions on T1-weighted images and as hypointense lesions on T2-weighted images. High signal intensity of the vitreous was observed in patients with vitritis and in those who were thought to have protein leaking into the vitreous as a result of impairment of the retinal-blood barrier.
The relative contrast between two tissues in a magnetic resonance (MR) image is shown to be quantifiable for any combination of pulse timing parameters, provided the intrinsic parameters are known. Based on multiple inversion-recovery and spin echo images, a region-of-interest T1, T2 and density analysis was conducted at 1.4T in selected patients with diagnosed neuropathology for various brain tissues. The resulting tissue parameters subsequently served to calculate the contrast-to-noise (C/N) ratio for typical tissue interfaces as a function of the operator-variable pulse timing parameters and the data were compared with the images. Although such calculations may be useful as a protocol selection aid, it is obvious that an optimized pulse protocol can only be established for a single tissue interface. The data also reveal that a T2-discriminating pulse sequence like Carr-Purcell-Meiboom-Gill with long repetition time, generally advocated as clinically most effective, may not always be ideal.