Contrast induced myocardial signal reduction: effect of lanthanide chelates on ultra high speed MR images.

The myocardial MR signal reduction associated with an intravenous bolus of Gd-DTPA and Dy-DTPA was studied in a canine model. Imaging was performed with a high speed echo-planar type imaging system (Instascan, Advanced NMR Systems, Inc.). Gated spin-echo images were obtained with TE of 30 ms, which permits image acquisition in approximately 40 ms. The gated TR was dependent on the heart rate, with an average TR of 2.4 s. After 0.1 mmol/kg of contrast was injected, 70 images were acquired, which showed in an 80-image data set a reduction in myocardial signal with a gradual return to normal. After dipyridamole infusion, the signal loss was significantly more pronounced, and earlier than in the control data set. There was no significant difference between Gd-DTPA and Dy-DTPA in these imaging studies despite the theoretical prediction of better Dy signal reduction, possibly due to physiological variability during the course of a study or between studies. The cause of enhanced contrast effect after dipyridamole infusion is discussed, as is the basis for dipyridamole enhancement, and the possible role of contrast enhanced MR imaging in the detection of cardiac disease.

Functional cerebral imaging by susceptibility-contrast NMR.

In vivo measurement of cerebral physiology by dynamic contrast-enhanced NMR is demonstrated. Time-resolved images of the cerebral transit of paramagnetic contrast agent were acquired using a new ultrafast NMR imaging technique and a novel mechanism of image contrast based on microscopic changes in tissue magnetic susceptibility. Global hypercapnia in dogs was used to establish the relationship between susceptibility-induced signal change and brain blood volume, and the response of gray and white matter to this microvascular stimulus was measured.

Forty-millisecond MR imaging of the abdomen at 2.0 T.

The ability of an ultrafast magnetic resonance (MR) imaging technique to provide abdominal MR images free of motion artifacts was studied. Individual T2-weighted transverse MR images were acquired in as little as 40 msec on a whole-body system operating at 2.0 T. Clinical evaluation was undertaken with fat-suppressed images in which only protons of water molecules contributed to image signal intensity. The ultrafast MR images were compared with conventional MR images obtained at 0.6 T. In 22 patients and two healthy volunteers, ultrafast MR images were of diagnostic quality and free of motion artifacts. Images obtained at an echo time (TE) of 30 msec (imaging time, 40 msec) had liver signal-to-noise ratios of 56.3 +/- 22.6 (n = 19). Because of a smaller data matrix, ultrafast MR images had soft-tissue interfaces that were less sharp than those of the highest-quality conventional MR images in which no motion artifacts were present. However, ultrafast MR images demonstrated high T2-dependent soft-tissue contrast, and pathologic and normal anatomies were readily detected with both imaging techniques. This ultrafast imaging technique has significant promise in whole-body MR imaging, in which motion artifacts often degrade image quality.

Instant images of the body by magnetic resonance.

Proton magnetic resonance (MR) body images of the normal, adult human which have total scan times of typically only 40 ms per image are presented. There is no loss of spatial or contrast resolution due to motional blurring or ghosting; rather, movie loops of multiple 40-ms images directly demonstrate normal respiratory and peristaltic motion. Manifestation of "traditional" relaxation time contrast is demonstrated for a variety of spin echo (TE) and image repetition (TR) times. The images, obtained at 2.0 T on a new high-speed MR system, have a signal-to-noise ratio for muscle of approximately 30:1 (TE = 30 ms) for a 4.7-mm slice thickness (voxel size = 0.08 cm3). In a study presented as an example, 140 images covering the body from diaphragm to pelvis were all obtained within approximately 10 min. This method may help improve the efficacy of MR body imaging in general, and may play a role in applications which require high temporal resolution.

Instant images of the human heart using a new, whole-body MR imaging system.

An extremely rapid MR imaging technique is described, and its use on a new 2.0-T high-speed MR system is demonstrated. This implementation permits complete filling of the two-dimensional spatial-frequency domain (k-space) within an acquisition window of 26 msec. With this acquisition window placed under the spin-echo signal envelope generated by a 90-180 degree pulse pair, the image contrast is the same as that of a conventional spin-echo pulse sequence. Resultant proton images have a motion-independent voxel resolution of 0.08 cm3 and a signal-to-noise ratio for cardiac muscle of approximately 30:1 (for TE = 30 msec) with no signal averaging. The pulse sequence yields images that are chemical shift-resolved. The total proton density distribution is optionally presented with lipid and water signals displayed in two different colors. Cardiac function is observed by displaying multiple images, acquired at different times in successive cardiac periods, in a cyclic movie format. Such motion pictures are obtained within a single period of suspended respiration, thereby assuring freedom from respiratory related motion artifacts. As preliminary examples, we present MR images of the normal adult human heart that have total acquisition times of only 40 msec/image and that show the major cardiac anatomy. Frames from movie loops show contraction of cardiac chambers and left ventricular wall thickening. The extremely rapid acquisition time of this technique suggests that it may hold promise for the routine and cost-effective evaluation of cardiac anatomy and function.

Proton chemical shift imaging: an evaluation of its clinical potential using an in vivo fatty liver model.

The potential laboratory and clinical utility of proton chemical shift imaging (PCSI) was evaluated by studying fatty liver change in rats, which offered a simple animal model for tissue lipid buildup. There was excellent correlation between lipid group signal intensities from in vivo PCSI studies and liver triglyceride levels obtained from in vitro measurements (R = 0.97). The in vivo T1 relaxation time measurements in fatty liver tissue demonstrated two distinct populations of nonexchanging protons. We explain the reason for the lack of sensitivity in conventional magnetic resonance (MR) imaging studies of fatty liver change and discuss the implications of our findings for MR imaging studies of other tissues. PCSI promises improved diagnostic sensitivity and specificity in studying a wide range of human pathologic conditions.

Nuclear magnetic resonance imaging in evaluation of clinical stroke syndromes.

Proton nuclear magnetic resonance (NMR) imaging was performed on normal volunteers and patients with various types of clinical strokes. True three-dimensional volumetric data were obtained for subsequent reconstruction of images at various orientations, including transverse, coronal, and sagittal, and for specific matching to x-ray computed tomographic planes. A variety of radiofrequency pulse sequences was used to generate images weighted by the NMR parameters spin density (p) and spin-lattice (T1) and spin-spin (T2) relaxation times. Quantitative T1 data and qualitative T2 information were obtained from appropriate sequences. Proton NMR images embodying T1 information provided adequate delineation of normal intracranial anatomical structure and a high level of gray matter-white matter contrast; high lesion detection sensitivity in various clinical forms of stroke was seen in images providing both T1 and T2 information. T1 and T2 relaxation times were prolonged in embolic, thrombotic, and watershed infarctions as well as in lacunae; shortening of T1 and T2 was found in hematomas. Serpiginous vessels in an arteriovenous malformation were seen that had particular pulse sequences.

Central pontine myelinolysis: demonstration by nuclear magnetic resonance.

An alcoholic, hyponatremic woman developed central pontine myelinolysis (CPM) and improved from a decerebrate, comatose state to alertness and full ambulation. NMR, using inversion-recovery and spin-echo pulse sequences, was performed sequentially from 4 weeks to 8 months after onset of symptoms and revealed a well-defined lesion with prolonged relaxation times. The lesion was anatomically consistent with CPM and was initially also visualized by CT. NMR showed no definite temporal change in the qualitative appearance of the lesion until the 8-month scan; however, quantitatively, a reduction of relaxation times was noted with each serial study.

Spin lattice relaxation time measurements in two-dimensional nuclear magnetic resonance imaging: corrections for plane selection and pulse sequence.

Accurate determination of relaxation times has become increasingly important in efforts to determine the diagnostic specificity of nuclear magnetic resonance (NMR) imaging. Techniques used in NMR imaging, not routinely employed in conventional NMR spectroscopy, can significantly affect the resulting relaxation time determinations. For the saturation recovery (SR) approach of T1 measurement used in our laboratory, these include selective excitation to define the image plane and magnetization refocusing for NMR signal acquisition. Computer modeling of the Bloch equations shows that errors well over 50% can be made in image derived T1 measurements if the conventional SR relation between signal intensity and the 90-90 degrees interpulse delay, tr, is used. However, corrected expressions can be derived for the actual pulse and gradient sequence used by our imaging system, and phantom data acquired in imaging experiments have verified the validity of these equations. This allows for the correction of T1 data to compensate for systematic bias introduced during imaging procedures and suggests a means whereby errors introduced by radio frequency inhomogeneities across the imaging volume can be reduced.

MR imaging of pituitary adenomas using a prototype resistive magnet: preliminary assessment.

Magnetic resonance (MR) images were obtained with a prototype resistive magnet system in 10 patients, all of whom had been shown to have pituitary tumors by enhanced high-resolution computed tomography (CT). Histologic verification was obtained in eight cases. Inversion-recovery (IR) T1-weighted images revealed the tumor in six of nine cases; saturation-recovery (SR) images with less T1 weighting identified seven of nine tumors; Carr-Purcell-Meiboom-Gill (CPMG) spin-echo T2-weighted images revealed two of four tumors. MR images failed to demonstrate three microadenomas: 5 X 5 X 8 mm, 6 X 6 X 6 mm, and one less than 5 mm in estimated size. In the last pretreatment study, CT had demonstrated a 13 mm maximum diameter adenoma. Repeat CT at the time of MR imaging also showed a partially empty sella and did not resolve the residual adenoma. The larger adenomas were identified readily by MR imaging, which, unlike CT, suggested old tumor hemorrhage in two cases, which was confirmed at surgery and histologic examination. MR and CT images were also compared for relative effectiveness in identifying important perisellar structures.

Nuclear magnetic resonance: in vivo proton chemical shift imaging. Work in progress.

Using a three-dimensional Fourier transform approach, proton nuclear magnetic resonance (NMR) chemical shift images have been obtained in vivo for the first time. At a proton resonance frequency of 61.5 MHz, chemical shift-resolved images of simple phantoms indicate that a spectral resolution of 0.7 parts per million (ppm) is readily achievable at all locations within the image matrix, even when using a magnet not specifically designed for chemical shift spectroscopy. In vivo images of the human forearm and of a cat head yield separable signals from water and lipid protons. However, using simple radiofrequency pulse sequences, our data show that relatively little signal originates from membrane lipids (e.g., myelin) in the brain. The measurement of magnetic susceptibility using this technique is also demonstrated. While helping to elucidate the genesis of the NMR response in complex biological systems, this methodology also has potential applications in medical diagnosis. The technique is also applicable to the chemical shift imaging of other nuclei; for example, phosphorus (P-31).

Instrumentation for nuclear magnetic resonance imaging.

The purpose of this article is to provide an overview of the instrumentation requirements of a nuclear magnetic resonance (NMR) imaging system. For the most part only concepts or generic descriptions of the system components are given, and it must be remembered that specific design criteria will be influenced by, among other things, the particular imaging methods to be used, and the proposed applications.

NMR imaging of leg tumors.

NMR images of 8 patients with neoplasms of the legs were obtained. Volumetric and/or planar NMR data were acquired using a saturation recovery (SR) approach, incorporating magnetization refocusing. NMR images revealed tumors in all patients and correlated well with the extent seen on CT. SR images with a short interpulse delay (tau) demonstrated a significant decrease in signal intensity (SI) in histologically normal fat (n = 4) and marrow (n = 1) adjacent to tumors, consistent with a prolonged T1. At certain values of tau, tumors on SR images could not be differentiated from normal muscle (tau = 200 msec.) and marrow (tau = 2,100 msec.) by SI alone. Using this sequence with fixed signal refocusing parameters, images representing several values of tau may be required to distinguish tumors from normal structures.

Nuclear magnetic resonance (NMR) imaging of tumors in the posterior fossa.

Nuclear magnetic resonance (NMR) images were obtained in 12 patients with mass lesions in the posterior fossa and the results compared with X-ray computed tomography (CT). Inversion recovery T1-weighted images demonstrated abnormalities in six of six intrinsic lesions and three of six extrinsic lesions. Spin echo T2-weighted images demonstrated abnormalities in two of two intrinsic lesions and four of five extrinsic lesions. Saturation recovery T1-weighted images were normal in two of two intrinsic lesions and two of four extrinsic lesions. Overall, NMR detected 11 of 12 lesions. Two intrinsic tumors detected by NMR were not detected by CT evaluation. Two small extrinsic tumors required CT gas cisternography for detection.

Preliminary clinical results of proton (1H) imaging of cranial neoplasms: in vivo measurements of T1 and mobile proton density.

Proton nuclear magnetic resonance (NMR) images reflecting T1 relaxation time and approximating proton density were acquired and used to generate T1 rate (1/T1) maps. By region-of-interest selection, measurements of T1 relaxation time were made from discrete volumes of the imaging plane. Such techniques were applied to the study of human cranial neoplasia and associated conditions of differential diagnostic importance (e.g., postoperative changes, radiation necrosis). Inversion-recovery NMR images exhibit a high lesion-detection sensitivity. In all patients, the specificity of NMR imaging is low since all abnormal areas appear as lesions darker than surrounding normal brain, reflecting a decreased proton density, prolonged T1 values, or both. T1 relaxation times are prolonged within neoplastic foci; however, absolute T1 values overlap with those found in other lesions.