Flow effects in multislice, spin-echo magnetic resonance imaging. Model, experimental verification, and clinical examples

Flowing blood is responsible for a number of complex effects on clinical magnetic resonance (MR) images. To help elucidate these effects, a computer model of a conventional multislice spin-echo pulse sequence was developed. Using TR, TE, and direction of slice acquisition, the model calculates and plots a profile of MR signal intensity vs. z-axis velocity. The model predicts complex profiles with multiple segments of MR signal loss depending on TR, TE, direction of flow, sequence and timing of slice excitation, and slice location relative to adjacent slices. Model predictions were verified by imaging a bulk-flow phantom, consisting of a rotating cylinder filled with a manganese chloride solution with T1 = 840 msec and characterized by a velocity-gradient resolution of 0.23 cm/sec/pixel. In conventional spin-echo MRI of medium and large vessels using body coils, in which the velocity gradients exceed 2-5 cm/sec/pixel, most of the flow artifacts are averaged and are difficult to appreciate. However, bright crescents or rings of MR signal occasionally are seen in the inferior vena cava and portal vein, which the model is invoked to explain. The bulk-flow phantom will find use as a tool for calibrating flow-sensitive pulse sequences when these become widely available.     

Visualization of swallowing using real-time TrueFISP MR fluoroscopy

The aim of this study was to evaluate the ability of different real-time true fast imaging with steady precession (TrueFISP) sequences regarding their ability to depict the swallowing process and delineate oropharyngeal pathologies in patients with dysphagia. Real-time TrueFISP visualization of swallowing was performed in 8 volunteers and 6 patients with dysphagia using a 1.5 T scanner (Magnetom Sonata, Siemens, Erlangen Germany) equipped with high-performance gradients (amplitude 40 mT/m). Image quality of four different real-time TrueFISP sequences (TR 2.2-3.0 ms, TE 1.1-1.5 ms, matrix 63 x 128-135 x 256, field of view 250 mm(2), acquisition time per image 139-405 ms) was evaluated. Water, yoghurt, and semolina pudding were assessed as oral contrast agents. Functional exploration of the oropharyngeal apparatus was best possible using the fastest real-time TrueFISP sequence (TR 2.2 ms, TE 1.1 ms, matrix 63 x 128). Increased acquisition time resulted in blurring of anatomical structures. As the image contrast of TrueFISP sequences depends on T2/T1 properties, all tested foodstuff were well suited as oral contrast agents, but image quality was best using semolina pudding. Real-time visualization of swallowing is possible using real-time TrueFISP sequences in conjunction with oral contrast agents. For the functional exploration of swallowing high temporal resolution is more crucial than spatial resolution.     

Quantitative analysis of sodium fast and slow component in in vivo human brain tissue using MR Na image]

In vivo sodium concentrations in the normal brain tissue and a tumorous tissue were analyzed using MR Na image. The nuclear magnetic resonance enabled us to divide the signal from sodium in the living tissue into 2 parts based on the differences of T2 value. Those are fast component having the T2 value of less than 5 msec and slow component of 15-40 msec. We investigated the effect of macromolecules on T2 value of sodium image using polyvinyl alcohol (PVA) powder. MR Na image was taken with the parameters of TR/TD, 110 ms/1.9 ms (FID image) and TR/TE, 110 ms/20 ms (SE image). Saline solution showed high intensity on both FID image and SE image. Saline solution added PVA (PVA phantom) also showed high intensity on FID image, whereas the signal intensity of PVA phantom in SE image extinguished. To know the relation between the signal intensity and sodium concentration, sodium concentration--signal intensity curve was obtained using phantoms with various sodium concentrations (0.05-1.0%). This curve showed a direct proportion between sodium concentration and signal intensity on Na image. We measured further the sodium concentrations of the human brain tissue. Sodium phantoms were arranged around the heads and the MR Na images of the normal brains from 3 volunteers and a patient with a brain tumor (meningioma) were taken. The sodium concentrations of occipital lobe, basal ganglia and the tumorous tissue were calculated using the sodium concentration--signal intensity curve obtained from the phantoms arranged around the heads.(ABSTRACT TRUNCATED AT 250 WORDS)     

Flow velocity of the cortical vein and its effect on functional brain MRI at 1.5t: Preliminary results by Cine-MR venography

The purpose of this study is to demonstrate the effect of altering flow velocity of cerebral cortical veins as the source of the signal change observed in functional magnetic resonance imaging (fMRI) of the brain. 10 healthy volunteers were examined after instructions in self-paced hand grasping. Experiments were performed using a 1.5-Tesla whole body MR scanner with a conventional two-dimensional gradient echo sequence (TR/TE/flip angle 400/60/40, first order flow rephased, reduced band width 8 Hz/pixel). Flow velocity measurements were performed for the cortical veins which corresponded to the activated areas depicted on fMRI. Velocity was estimated from the cine-MR venography (cine-MRV) with a tagging technique. Flow phantom studies were performed to delineate the effect of flow velocity differences upon the subtraction images of fMRI. The cine-MRV revealed increased flow velocity of the cortical veins during activation in seven volunteers, with a mean velocity difference of 15 mm/sec. Flow phantom studies suggested that the increased flow velocity may result in changes of the flow signal profile due to oblique flow displacement. Subtraction of the two images with different flow profiles produces flow signal enhancement. Increased flow velocity of the cortical veins during the activation is an important factor which contributes to the signal of fMRI.     

MR imaging of intracranial hemorrhage in neonates and infants at 2.35 Tesla

Summary The variations of the relative signal intensity and the time dependent changing contrast of intracranial hemorrhages on high-field spin-echo magnetic resonance images (MRI) were studied in 28 pediatric patients. For T1-weighted images, a repetition time (TR) of 500 ms and an echo time (TE) of 30 or 23 ms was used. The corresponding times for T2-weighted images were TR 3000 ms and TE 120 ms. Intracranial hematomas, less than 3 days old, were iso- to mildly hypointense on short TR/TE scans and markedly hypointense on long TR/TE scans (acute stage). In the following four days the signal of the hematomas became hyperintense on short TR/TE scans, beginning in the periphery and proceeding towards the center. On long TR/TE scans the signal remained markedly hypointense (early subacute stage). 7–14 days old hematomas were of high signal intensity on short TR/TE scans. On long TR/TE scans they appeared hypointense in the center and hyperintense in the periphery (late subacute stage). By the end of the second week the hematomas were of high signal intensity on all pulse sequences (chronic stage). Chronic hematomas were surrounded by a parenchymal rim of hypointensity on long TR/TE scans. 28 neonates and infants (with 11 follow-up examinations) of 31.5–70.6 weeks postconceptional age (PCA), with an intracranial hemorrhage were examined. The etiologies of the hemorrhages were: asphyxia (17 cases), brain infarct (2), thrombocytopenia (1), clotting disorder (1) and unknown origin (7). The aim of this study was to describe the appearance of intracranial hemorrhages inneonates and infants with MRI at2.35 Tesla using spine-cho sequences.     

Real-time cardiac cine imaging with SPIDER: steady-state projection imaging with dynamic echo-train readout.

Steady-state projection imaging with dynamic echo-train readout (SPIDER) is a multiecho radial k-space trajectory TrueFISP sequence developed for real-time cine imaging of the heart. This new pulse sequence combines the superior SNR and blood-to-myocardium contrast of TrueFISP with the increased scan time efficiency of EPI and undersampled projection reconstruction. SPIDER sequence RF repetition time (TR) was minimized by limiting the echo-train to a length of three while acquiring the first and third echoes asymmetrically. A temporal resolution of 45 ms was achieved with TR/TE1/TE2/TE3 of 3.24/0.6/1.6/2.6 ms and a factor of 2 view sharing scheme. Phantom experiments showed little difference between the weighting of the signals acquired at each of the echo times but did show considerable off-resonance modulation between them. In vivo experiments demonstrated the feasibility of using the SPIDER sequence for real-time imaging in the cardiac short axis orientation.     

The magnetic resonance imaging appearance at 1.5 Tesla of cartilaginous tumors involving the epiphysis

Three cases of lytic, calcified epiphyseal lesions with plain film and computed tomography features suggestive of chondroblastoma were imaged by magnetic resonance imaging. Histopathologic correlation was obtained in each case. Two cases of chondroblastoma showed low signal intensity on both short (TR600/TE20ms) and long (TR2500/TE80ms) spin echo (SE) images. The third case, a clear cell chondrosarcoma, demonstrated increased signal intensity on moderately T2 weighted (TR2500/TE40ms) images. These findings suggest that magnetic resonance imaging may be helpful in distinguishing these lesions.     

Elimination of errors caused by first-order aliasing in velocity encoded cine-MR measurements of postoperative jets after aortic coarctation: in vitro and in vivo validation

The aim of this study was to evaluate a velocity-encoded cine-MR (VEC-MR) sequence in measuring flow velocities up to two times the velocity encoding value (VENC) in a flow phantom and to validate the method for assessing poststenotic jet velocities in postoperative patients after aortic coarctation. In vitro, a flow phantom was used (0.5T; TR/TE: 51/8 ms, flip angle=30 degrees, FOV=280 mm, 128x256 matrix VENC 40 or 80 cm/s). On binary images, maximum flow velocities (V(max)) were calculated with a region of interest (ROI, 8 pixels). With aliasing, V(max) was calculated by VENC+(V(aliasing)). In vivo, 16 postoperative patients after aortic coarctation underwent double-oblique VEC-MR imaging through the aortic arch (ECG triggering, 16 phases/RR, TR=600-800 ms, flow-encoding cranio-caudal, VENC=2 m/s). Peak systolic velocities were measured and transthoracic Doppler echocardiography (TTDE) was performed. In vitro, there were excellent correlations for MR velocity measurements with and without aliasing ( r=0.99) and for true and MR-derived flow velocities ( r=0.99). In vivo, there was good correlation between VEC-MR and TTDE-assessed V(max) values in the aorta at the former coarctation site ( r=0.90, n=16). Aliasing occurred in 13 patients. VEC-MR is a useful modality for assessing jet velocities in the follow-up of patients after aortic coarctation. Despite of aliasing, accurate velocity measurements up to two times VENC are possible using binary images.     

MR imaging evaluation of plica synoviallis mediopatellaris of the knee joint]

To evaluate the diagnostic ability of MR imaging for plica synoviallis mediopatellaris (PSM), we retrospectively reviewed the MR imaging findings of patellofemoral space in 20 knee joints of 11 patients. In all 20 knee joints, arthroscopy and MR imaging were available. MR imaging was performed with a 1.5 Tesla Magnetom (Siemens) using a round surface coil. Pulse sequences were SE (TR 600 ms/TE 26 ms), SE (TR 200 ms/TE 26, 70 ms) and FLASH (TR 450 ms/TE 15 ms/FA 90 degrees). In six of the 20 knees with PSM proved by arthroscopy, a low intensity band was shown above the medial condyle of the femur on both T1- and T2-weighted MR images, and on FLASH images this band was shown as intermediate intensity. In the other 14 knees with no PSM observed by arthroscopy, the low intensity band was not shown on MR imaging. In all 20 knees, a similar low intensity band was shown about 1 cm cranial to the medial condyle of the femur. This should not be diagnosed as PSM. The low intensity band seen on T1- and T2-weighted MR images and its anatomical relation to the medial condyle are important in diagnosing PSM.     

MR imaging of peripheral cholangiocarcinoma.

A prospective study was performed to compare MR spin-echo (SE) sequences [repetition time/echo time (TR/TE) 2,000/80, 500/44 ms], unenhanced CT, and rapid intravenous contrast enhanced CT in eight consecutive patients with peripheral cholangiocarcinoma. All the tumors (ranging from 5 to 9.6 cm in size) were detected with all four techniques. Tumor contrast, however, was qualitatively greatest on long TR/TE SE images. With long TR/TE SE images, tumors were demonstrated as well-demarcated homogeneous regions of high signal intensity, and the anatomic relations between tumors and intrahepatic blood vessels were easily perceived. Detection of small intrahepatic metastatic foci was best on long TR/TE images. Tumor invasion of the portal vein's branches was also best seen on long TR/TE SE images. These results indicate that long TR/TE SE sequence is the most effective initial screening method in demonstrating the presence and determining resectability of peripheral cholangiocarcinoma.     

Clinical significance of the intranuclear cleft of the lumbar intervertebral disk on MRI

MR studies of the lumbar spine in 111 patients were analyzed at 469 disks to assess the prevalence of intranuclear cleft (INC) in the lumbar intervertebral disk. MR studies were performed on either 0.1-tesla (T) magnet (69 patients) or 0.22-T magnet (42 patients). The pulse sequences reviewed were saturation recovery (SR; TR = 0.5 sec), short TR, TE spin echo (S-SE; TR = 0.5 sec, TE = 34 msec) and long TR, TE spin echo (L-SE; TR = 1.5 sec, TE = 68,80 msec). All study were done in a sagittal plane with 10 mm slice thickness. The conclusions were as follows: 1) On a 80 msec TE, 1.5 sec TR image, INCs were detected in more than 80% of disks in patients over 30 years old but in only 13.3% of disks in patients under 20 years old. 2) In both imaging system, L-SE showed INCs more frequently than SR and S-SE. 3) INCs were less frequently demonstrated in the disk with decreased signal intensity on 0.1-T magnet as compared with 0.22-T magnet. 4) On SR and S-SE, there is an increase in the prevalence of INC in the disk with decreased signal intensity. We suggest that the INC will be a good landmark of the pathological process of the lumbar disk, such as degeneration.     

Dural sinus occlusion: evaluation with phase-sensitive gradient-echo MR imaging

The purpose of this study was to evaluate the usefulness of limited-flip-angle, phase-sensitive velocity imaging with gradient-recalled-echo (VIGRE) MR when combined with spin-echo MR in the diagnosis of dural sinus thrombosis. The VIGRE sequence consists of a rapid single-slice acquisition, 50/15/2 (TR/TE/excitations), and 30 degrees flip angle. At each slice position, a total of four images were reconstructed; these consisted of one magnitude image and three images sensitive to proton motion in each orthogonal direction. The flow direction and flow velocity (cm/sec) were obtained from each of the phase images, and results were correlated with data obtained from a phantom experiment. In normal controls, dural sinus velocities ranged from a mean of 9.9 to 14.4 cm/sec for the transverse and superior sagittal sinuses, respectively. Three patients with proved dural sinus occlusion were studied with spin-echo images at 1.5 T. Three-dimensional time-of-flight MR angiography was also performed in one patient. The presence of dural sinus occlusion was determined by the lack of flow void on the spin-echo images, the absence of phase shift on the VIGRE study, and the presence of retrograde flow on the phase image in the sinus proximal to the occluded segment. Time-of-flight angiography overestimated the extent of the thrombosis caused by spin saturation. Follow-up VIGRE studies detected the formation of collateral flow in one patient and recanalization with the establishment of normal antegrade sinus flow in the other. We conclude that phase-sensitive MR imaging is helpful in establishing the diagnosis and extent of dural sinus occlusion.(ABSTRACT TRUNCATED AT 250 WORDS)     

Morphology and hemodynamics in dissection of the descending aorta. Assessment with MR imaging

Purpose: To obtain morphologic and functional information in patients with dissection of the descending aorta using contrast-enhanced MR angiography (MRA) and MR blood flow quantification of the true and false lumina.Material and Methods: Fourteen patients were studied prospectively using a 1.5 T unit. MRA was performed with a 3D FISP sequence (TR/TE/flip angle 4.7/1.9 ms/30°) after injection of 0.2 mmol Gd-DTPA per kg b.w. Flow quantification with phase velocity mapping was done at the level of the diaphragm using a 2D FLASH technique (TR/TE/flip angle 28/6.5 ms/30°) with an average temporal resolution of 23 frames per cardiac cycle (34 ms). A spectral broadening index was applied to quantify the amount of flow irregularity within both channels of the aorta. Extension of the dissection and involvement of the major branch vessels were analyzed.Results: The mean flow volume per minute was 1982 ml (SD 1083 ml) in the true and 1052 ml (SD 763 ml) in the false lumen. Average peak-velocities were 98 cm/s (SD 33 cm/s) in the true channel and 47 cm/s (SD 26 cm/s) in the false channel. Ten patients had bidirectional flow in the false lumen with a reflux volume ranging between 6.8% and 98%. Only 1 patient presented with bidirectional flow in the true lumen (reflux volume 15%). A significantly higher degree of flow irregularity was found in the false lumen compared with the true channel.Conclusion: Different hemodynamic patterns were found in aortic dissection. Their prognostic value and the impact on therapy, specifically percutaneous interventional procedures, have to be further studied.     

Interindividual reproducibility of glutamate quantification using 1.5-T proton magnetic resonance spectroscopy.

The goal of this study was to measure the interindividual reproducibility of glutamate quantification in 1.5-T (1)H MRS of human brains. To determine the effective echo time (TE) for glutamate quantification, spectra from a phantom and 12 participants were obtained with TE = 30, 35, 40, and 144 ms (repetition time (TR) = 2000 ms and volume of interest = 4 cm(3)). The average Cramer-Rao lower bounds for glutamate quantification using LCModel was lowest in two experiments when TE = 40 ms.Twenty-one subjects participated in experiments that measured interindividual reproducibility of glutamate quantification. Spectra were acquired with TR = 6000 ms and TE = 40 ms. Results showed that the coefficients of variance were 11.0 and 13.1% in the anterior cingulate cortex and insula, respectively. This suggests that glutamate can be reproducibly measured from 1.5-T (1)H MRS with long TR, effective TE, and the LCModel.     

Delineation of simulated vascular stenosis with Gd-DTPA-enhanced 3D gradient echo MR angiography: an experimental study

PURPOSE: The purpose of this experimental study was to evaluate the influence of contrast material concentration and flow velocity on pulsatile flow in Gd-DTPA-enhanced 3D gradient echo MR angiographic sequence. METHOD: In vivo flow experiments were performed in Plexiglas phantoms with artificial stenosis (50% stenotic ratio and 20 mm stenotic length) attached to a cardiac pump that generated physiological pulsatile flow similar to that of the bloodstream in a closed circuit. We used a steady-state gradient echo sequence with different TEs (6, 3, and 1.4 ms). A TR of 15 ms was used for all parameters. The concentration of Gd-DTPA varied from 0 to 2.0 mmol/L and flow velocities from 25 to 80 cm/s. We measured the degree of stenosis and length of stenosis in comparison with the actual values. RESULTS: The degree and length of stenosis on 3D gradient echo MR angiographic images were markedly influenced by the velocity of the flow and concentration of Gd-DTPA. The degree of stenosis was overestimated when the flow was fast or when the concentration of Gd-DTPA was low. When the concentration of Gd-DTPA was low, stenosis was elongated. These effects were less prominent on short TE (1.4 ms) sequence. CONCLUSION: The stenotic lesions were markedly overestimated on MR angiographic images obtained with Gd-DTPA-enhanced fast 3D gradient echo sequence. Spin dephasing can be compensated for almost entirely by a high concentration of Gd-DTPA and/or a short TE sequence.     

Report on the ECR2003 (European Congress of Radiology): non-contrast intracranial MR venography with 3D-phase contrast imaging: optimization of presaturation pulse and velocity encoding

BACKGROUND: Intracranial MR venography is useful for the diagnosis of dural sinus thrombosis and the preoperative assessment of sinus patency encased by tumors. Recently, contrast-enhanced MR venography has been applied for suspected dural sinus occlusion in a shorter time. However, it has some disadvantage for the evaluation of hypervascularized enhancing thrombus mimicking flow in chronic sinus thrombosis. So far, we have evaluated optimal imaging technique and slice orientation and have shown that sagittal three-dimensional (3D) -phase contrast (PC) imaging is the most suitable for the non-contrast intracranial MR venography. PURPOSE:To assess the optimal presaturation pulse (SAT) and velocity encoding (VENC) for the non-contrast intracranial 3D-PC MR venography. METHODS AND MATERIALS: Firstly, we performed phantom experiment to assess the best SAT thickness using arterial presaturation. Second, MR imaging was performed in 7 healthy volunteers to measure the dural sinus flow velocity using a 1.5 T MR. Third, 3D-PC MR venography was performed with a VENC settings at 10, 15, 20 and 30 cm/sec for healthy volunteers. All data were displayed as maximum intensity projection images and three neuroradiologists assessed the visibility of the dural sinuses and the cortical vein. RESULTS: The mean flow velocity of the dural sinuses was 6.3 cm/sec. The thickness of the best SAT was 100 mm. In the assessment of the visibility of the 3D-PC images, dural sinuses were adequately visualized at a VENC of 15 cm/sec. CONCLUSIONS: Non-contrast intracranial 3D-PC MR venography was optimized at 100mm thickness of SAT and a VENC of 15 cm/sec.     

Heating and safety of a new MR-compatible guidewire prototype versus a standard nitinol guidewire

Our purpose in this study was to examine heating of nitinol and polyetheretherketone (PEEK) guidewires during near-real-time MR imaging in an artificial vascular model an “aorta phantom”. The first 100 cm of the nitinol- and PEEK-based guidewires both 145 × 0.08 cm were immersed in a saline-filled aorta phantom. The probes of a fiber-optic thermometer were positioned at the tips of both wires. Balanced steady-state free precession (bSSFP) [TE 1.6 ms; TR 3.5 ms; flip angle (FA) 60°; field of view (FOV) 40 cm; matrix 256 × 256; specific absorption rate (SAR); 1.15 Watt (W)/kg] and spoiled gradient-echo (SPGR) (TE 1.8 ms; TR 60 ms; FA 60°; FOV 40 cm; matrix 256 × 256; SAR 1.15 W/kg) pulse sequences were acquired in a 1.5-T MR scanner with use of an 8-channel array coil. Temperatures were recorded while the phantom was placed centrally in the bore of a MR scanner and in an off-center position (x = 24 cm, y = ?5 cm, z = ?10/10 cm). The temperature of the nitinol guidewire increased by 0.3 °C (center) and 1.1 °C (off-center position) with use of the bSSFP and by 9.6 and 13 °C (off-center position) with use of the SPGR sequence. Only minor temperature changes up to a maximum of 0.4 °C were observed with the MR-compatible PEEK guidewire when any position or sequence was applied. The PEEK guidewire showed substantially lower heating as compared to the nitinol guidewire in near-real-time imaging sequences in a phantom.     

MR imaging of the dilated biliary tract

The magnetic resonance (MR) examinations of 18 patients with dilated bile ducts were reviewed retrospectively to determine the capability of MR to demonstrate biliary dilatation, assess MR appearance of the dilated biliary tract using spin-echo techniques, and define the optimal MR imaging parameters (repetition time [TR] and echo time [TE]) for its demonstration. On images with short TR (0.5 sec) and TE (28 msec), the dilated intrahepatic and intrapancreatic bile ducts usually had lower signal intensity compared with the surrounding liver or pancreas; on images with long TR (2.0 sec) and TE (56 msec), they had higher signal intensity. Because of the observed variation in percentage of contrast between dilated bile ducts and surrounding liver and pancreas, two imaging sequences are recommended to obtain reliable demonstration of dilated intrahepatic and intrapancreatic bile ducts. The dilated common bile duct at the level of the hepatic hilus is best seen with a short TR and TE.     

Contrast-enhanced MRA of the brain.

Most sequences for MR angiography (MRA) used today exploit the macroscopic motion of the blood to differentiate vessels from the stationary tissues. An alternative approach to inflow based MRA is contrast enhanced MRA, in which relaxation agents are used to selectively shorten the T1 of the blood below the T1 value of the stationary tissues. We have evaluated cerebral Gd enhanced MRA, comparing it with conventional angiography and noncontrast inflow based MRA. Contrast/enhanced MRAs were obtained at 1.0 T with a 3D FISP sequence with TR/TE/alpha: 35-40 ms, 7-11 ms/TE/25 degrees. Contrast enhancement was obtained by a biphasic injection of a double dose of Gd-DOTA (0.2 mmol/kg) during image acquisition. With the described technique the conspicuity of both cerebral arteries and veins is improved compared to nonenhanced inflow MRA.     

Anterior and posterior lobes of the pituitary gland: assessment by 1.5 T MR imaging

Pituitary glands of 60 normal volunteers (30 men 20-36 years old, and 30 women 18-42 years old) were studied by 1.5 T magnetic resonance (MR) imaging. The T1-weighted images (T1WI) [repetition time (TR) = 400 ms; echo time (TE) = 25 ms] were obtained in the coronal, sagittal, and axial planes. Proton density (PD)/T2-weighted images (PDWI/T2WI) (TR = 2,000 ms; TE = 25/100 ms) were obtained in the sagittal plane using 3 mm slice thickness. On T1WIs of all subjects the posterior part (PP) of the pituitary fossa showed the highest signal, which was indistinguishable from fatty tissue. This study reveals that this region of high signal intensity (PP) corresponds to the posterior lobe and not intrasellar fat because its shape, size, and position are compatible with the posterior lobe; its signal intensity differs from that of fatty tissue on PDWI and T2WI; the absence of an intrinsic chemical shift artifact (CSA) characteristic of fat; and due to CSA, a dorsum with fatty marrow is shifted relative to the PP (or may be made to merge with it). Regarding the differentiation of the two lobes of the pituitary gland on MR, the morphology of the anterior and posterior lobes was evaluated and great variation found. Appreciation of normal is particularly important in evaluating coronal images for small pituitary lesions.