Peter Reimer, MD #{149} Sanjay Saini, MD Thomas J. Brady, MD #{149} Mark S. Cohen,

Techniques Echo-Planar

could

ference

Index

be identified

in signal

atitis,

terms:

by their

C

intensity.

Magnetic

resonance

#{149} Magnetic

enhancement

770.291

(1).

The

two

principal

#{149} Peter

R. Mueller,

any

to gross

reasons

physiologic

and

the lack

trast

agent

many])

for

motion

of a suitable

this

second(2,3)

bowel

and

various Erlangen,

MR

con-

because

noise ratio abdominal restricted

and

low

liver

(14).

Since

these

MR

improved

nan MR images can variety of strategies

enhancement

bowel

contrast

with

on

aqueous

the

MGH-NMR

Massachusetts

General

RSNA,

1992

a section

was

(Signa;

GE Medical

with

Systems,

Center Hospital

and

Division and

Harvard

of Gastrointestinal Medical

Radiology, School,

32 Fruit

Department St. Boston,

[TR]

size

of 128

and

images

a 10-mm

in eight x 256

pa-

was

used

of 40 cm. These patients bowel lumen enhanceTIW

axial

of 10 mm

gap

section

times

T2W

fat-sup-

(15) were

obtained

thickness gap. pulse

(TIs)

and

Multisection sequences

selected

a invenwith

to approxi-

the null points of fat (TI = 100 msec), and pancreas (TI = 380 msec), and (TI = 800 msec) were used for TIW

oc) and

=

(14,16,17).

with

For

a TE

T2W

(26, 50, and “high-resolution” were

100

obtained

ac-

a multisection

SE pulse sequence was used. these SE images were acquired gle excitation pulse (TR = ) of TEs These

were pulse

of 26 msec

images,

Data for after a sinat a variety

msec) (14,16,17). echo-planar

with

“standard-resolution”

METHODS

performed

thickness

intersection

(64 x 128) (8,10,1 1).

imaging

(T1W) se-

time

imaging. Data for these images quired after a single excitation

ages

of

previ-

Ti-weighted with a pulse

(repetition

A matrix

mate liver spleen

(TR

or its

details

a matrix

im-

size

of

128 x 128, which, for a 40 x 20-cm field of view, yielded an in-plane resolution of 1.6 x 3.1 mm (14,16,17), and provided betten anatomical resolution than did the

Imaging

MR

Received June 27, 1991; revision requested July 24; revision Supported in part by grant CA Pol 48279 from the National development fund, and by the Society of Computed Body Deutsche Forschungsgemeinschaft (grant no. Re 758/1-2). (

with a 5-mm

echo-planar

described

To examine

182:175-179

features,

time [TE] msec], four excitaT2-weighted (T2W) SE images two excitations) were ob-

inversion

a i.5-T

MR

Mil-

resolution, was

, From

of 250/20

with

a

Insta-

the conof the Signa

described

axial images

5-mm intersection sion recovery (IR)

with acquisi-

The alter

technical

been

Echo-planar

agent.

AND

have

The

echo-planar

(14).

pressed

echo-pla-

an

system

size.

at a field of view did not undergo ment.

imaging

be obtained for image

bore

tients.

ana-

tion, we compared different echoplanar techniques for MR imaging of the pancreas. In addition, we exammed the potential utility of bowel lumen

comfort

its patient

tamed

or to dynamic

of conventional

imager,

and

motion studies of the bowel (8,11). Technical advances provide anatomic resolution and S/N values in echo-planar images comparable with those

did not capability

msec/echo tions) and (2,500/50-100,

signal-to-

(S/N) (7,10-13). Reports applications have been to the

scan modification ventional imaging

quence

imaging

of suboptimal

resolution

(Instascan).

Conventional spin-echo (SE)

Mass) have been developed to overcome motion artifacts in MR imaging of the abdomen and chest (5-9). Diagnostic application of echo-planar imaging techniques has been limited, tomic

to provide

capability

ousby

techniques such as Instascan (Advanced NMR Systems, Wilmington,

however,

modified

imaging

this

fast (TurboGer-

echo-planar

waukee)

imaging

(4).

In recent years, FLASH [Siemens,

imager

ogy,

MD

Pancreas’

artifacts

MATERIALS

1992;

PhD

magnetic resonance (MR) imaging has been of limvalue in examining the pancreas

are phase-encoding

MR Radiology

MD,

ONVENTIONAL

ited

dif-

(MR), conresonance (MR), echo planar #{149} Magnetic resonance (MR), pulse sequences #{149} Magnetic resonance (MR), tissue characterization #{149} Pancreas, MR, 770.1214 Pancreas, neoplasms, 770.321, 770.34 #{149} Pancretrast

F. Hahn,

PhD

for High-Resolution MR Imaging ofthe

Recent technical advances in echoplanar magnetic resonance (MR) imaging prompted an investigation of these new techniques in pancreatic MR imaging and evaluation of bowel lumen enhancement with an aqueous bowel contrast agent. In 42 subjects (36 healthy, six with pancreatic disease), various Ti-weighted inversionrecovery and T2-weighted spin-echo fat-suppressed pulse sequences were assessed with an echo-planar technique implemented with a modified clinical MR imager. Single-excitation imaging (echo time, 26 msec) provided a higher (P < .05) signal-tonoise ratio than did conventional spin-echo and all other echo-planar techniques. In i3 (72%) of i8 healthy subjects who did not undergo administration of the contrast agent, the entire pancreas was distinguished from adjoining bowel. In all 18 subjects who underwent contrast-enhanced imaging, a significantly greater (P < .05) intraluminal signal intensity was apparent with all echoplanar pulse sequences and the entire pancreas was identified. In six patients with pancreatic disease, be-

sions

#{149} Peter

images used

the utility with

even

the “mosaic” with

the

images

in previous

T2W

reports

of echo-planar higher

anatomic

method SE pulse

(14,16,17) sequence

of RadiolMA

02114.

received August 8; accepted August 16. Institutes of Health, by the MGH-NMR Tomography. P.R. supported by the Address reprint requests to 5.5.

Abbreviations: C/N = contrast-to-noise IR = inversion recovery, SE = spin echo, signal-to-noise ratio, TE = echo time, TI version time, TR = repetition time, TIW weighted, T2W = T2-weighted.

ratio, S/N =

in-

=

TI-

=

175

Figure 1.

Echo-planar and conventional MR images of the pancreas with no bowel lumen enhancement. (a) T2W high-resolution (single excitation, oo/26) SE image (top) and maximum-resolution (two excitations, 6,000/26) SE image (bottom) in a fasting subject. The pancreas is visualized because signal from netnoperitoneal fat is suppressed. The pancreas has a higher S/N on the single-excitation image, while anatomic resolution is better tional T1W (250/20)

on the two-excitation image. Note (top) and T2W (2,500/50) (bottom)

to achieve a 128 x 256 data matrix and an in-plane resolution of 1.6 x 1.6 mm. This “maximum-resolution”

quired

technique

separate

two

doubled

the breath-hold

seconds.

TR, which

between

the

tions,

ne-

excitations,

which

time from 6 to 12

represents

the

time

two radio-frequency excitaset at 6 seconds to minimize TI

was

effects (14,16,17). SE T2W maximum-nesolution images were also obtained with TEs of 26, 50, and 100 msec. All images were acquired with a constant 62.5% k-space coverage to provide a TEminimum of 26 msec (TEminimum at 100% k-space

coverage

tiab survey

is 73 msec)

examination

high-resolution 26)

image

men

(14).

with

single-excitation covered

during

the

a 6-second

mi-

An

a 21-section T2W

entire

upper

breath

(c/

We

then restricted the pancreas

the cephabocaudal range for subsequent acquisitions.

The subjects

were

instructed

to hold

to

their

breath for the duration of each acquisition (6 seconds for the single-excitation and 12 seconds for the two-excitation technique)

to minimize spatial misregistration tween the multisection acquisitions.

Figure normal ment. c/50)

be-

2.

Echo-planar

(36

was performed

with

a normal

The

portion

bile duct

two

with

pancreatitis, two with pancreatic carcinoma, and two with penipancreatic lymphoma). Eighteen subjects underwent imaging without any bowel preparation and 24

subjects

six patients 176

(18

with

with

#{149} Radiology

a normal

pancreatic

pancreas,

disease),

arrow)

as a high-signal-intensity

all

after

corn-

solution

All subjects had

at computed

in the

a normal

tomography

pathologic with either

informed

of the

can be identified

of a contrast

biopsy

consent

pan-

(CT).

conditions

study was approved ies Committee at our ten

convenimages.

structure.

oral administration as described below. healthy study group

creatic firmed

(b) Comparison on echo-planar

tail

mon

(small

secretions. S/N than

i

ages.

intrapancreatic

gastric a lower

enhanceexcitation, before

oral administration 2. The pancreatic

were

Pan-

con-

or surgery.

The

by the Human Studinstitution, and writwas

obtained

from

all subjects. Lumen

Enhancement 450

Westbury,

mL

of Readi-Cat

NY)

was

orally 15-30 minutes before cording to the manufacturer’s

in 42 sub-

pancreas,

of the

from adjoining on postcontrast im-

(E-Z-Em,

MR imaging

due to native pancreas has

can be clearly distinguished bowel (large arrow) only

In 24 subjects,

jects

MR images

(top) and after (bottom) of 450 mL of Readi-Cat

Bowel

Subjects

gastric antrum also shown. The

pancreas after bowel lumen T2W high-resolution (single SE images of normal pancreas

creas

abdo-

hold.

the hypenintense SE images are

Readi-Cat

2 is a dilute

2

administered

imaging acinstructions.

suspension

of bar-

ium that is commonly used in CT for bowel opacification. This preparation was chosen because of its tonicity (which distends the bowel), well-known administration methodology, safety, wide availability, and predictable signal intensity be-

havior. trast

The

high and

high

material signal

water

was intensity

SE echo-planar

enhance

the

of bowel

loops.

content

intended to bowel MR

images

2%

barium

intnaluminal The

of this

con-

to contribute lumen

on

and

thus

signal

IR

intensity content

of

Readi-Cat 2 renders it opaque at CT and is not a significant contributor to the MR signab intensity. No adverse reactions to the imaging

procedure

material

were

or to the

reported

oral

by any

contrast

subjects.

January1992

fenentiation

(binary

creatic

head,

addition,

images

tively

for

the

structures, and by

decision)

body,

were

Figure 3. Echo-planar MR images of pancreatic carcinoma after bowel lumen enhancement. Two adjacent T2W high-resolution (single excitation, c/50) SE images (10-mm sections with 5-mm gap). The cancinoma (arrow) can be identified as a highsignal-intensity area compared with the normal tissue in the pancreatic head (C/N, 13.8 ± 1.1). Atrophy with fatty replacement is seen in the body and tail of the pancreas. The pancreatic duct is dilated and visible as a high-signal-intensity structure.

by the noise

surrounding of interest

T2W SE (2,500/50,

Pancreatic S/N*

intensity

excitation) Echo-planar

MR

± 4.6

10.6

±

tail.

Re-

the

lu-

measuring

bowel wall. Regions within individual disease were used contrast-to-

(ri

17.3

of the

enhancement was

test

SE (TE, 26 msec) SE (TE, 50 msec) SE (TE, 100 msec) ane expressed

as mean

lumen

with

protocol

well

tolerated

Analysis

the

with-

signal Fisher

experienced dominal MR were analyzed

182

visible

by

analyzed

independently (P.R.,

S.S.,

in the interpretation and CT images. qualitatively

#{149} Number

i

P.R.M.)

of abThe images for clean dif-

signal

Overall

enhancement.

tomic

appeared

signal

interfaces,

with sunno visible the singleAna-

however,

with

had

a CT scan-bike

image

quality

subjects, in fat sup-

apparent

in the

The intrapancreatic common bile duct

was

substruc-

be identified

com-

and distinguished from adbowel with each imaging (Fig 1). This ease in identifiis best explained by intrinsic secretions that appeared hy-

an abnormal

intensity

pancreas.-

relative

to that

of nor-

mal pancreas (Fig 3). Tumor tissue and areas of pancreatitis demonstrated higher signal intensity than pancreas. The also identified

changes

common

bile

diseased by means

such duct,

of

as a di-

dilated

pan-

creatic duct, mesenteric stranding, pancreatic contour deformity. Quantitative With

high-intenwater con-

ture, and portions of the pancreatic duct could be identified as a highsignal-intensity structure in eight (22%) of 36 healthy subjects. In 13 (72%) of 18 subjects without bowel lumen enhancement, the entire could

of

intraluminab and

lated

pancreas-In

in 32 (89%) of 36 healthy as a high-signal-intensity

pletely joining technique cation bowel

and

In both groups (without bowel lumen enhancement), rounding bowel produced motion artifacts with either or two-excitation technique.

morphologic

Analysis

usually (high

images

especially

pancreas

radiologists

and lumen

these

pression

jects

was

distention

did normal areas were

a normal

larger subjects. segment of the

± I standand

bowel

conspicu-

because

and

with

of bowel

varied somewhat among with reduced efficiency

tissue.

patients.

were

tent),

18.1 ± 5.4 12.2 ± 3.4 6.2 ± 2.7

deviation for normal pancreatic t Highest S/N (P < .05)

with

bowel

Preliminary studies in six patients with pancreatic disease and bowel enhancement demonstrated that diseased areas or lesions were seen readily on the basis of their different

analyzed

appearance.

T2W

of surrounding

lumen

the

of back-

(18).

suppression) sity bowel

29.4 ± 96t 18.5 ± 6.4 8.3 ± 3.0

ity

Bowel

improved

deviation

pancreas

Image

< .05).

sharper on single-excitation images than on two-excitation images when the longer breath holds for two-excitation images could not be sustained.

all cases, the echo-planar technique produced diagnostic-quality images free of motion artifacts. With bow-intensity retropenitoneal fat (due to fat

± 7.7

(P

enhancement

signal intensity

Analysis

Subjects

2.9 ± 1.5 13.2 ± 4.5

significant

[mean mean signal =

noise).

Qualitative

36)

=

penintense on IR and SE echo-planar MR images. In the other five subjects, the pancreatic tail could not be distinguished clearly from collapsed or airfilled adjacent bowel. In all subjects with bowel lumen enhancement, the entire pancreas (head, body, and tail) was identified (Fig 2). The difference between the bowel-enhanced and bowel-unenhanced study groups was statistically

Subjects

4.0

TIW

Two-excitation

Volume

-

of lesion

Statistical

one

JR (TI, 100 msec) IR (TI, 380 msec) IR (TI, 800 msec) Single-excitation T2W SE (TE, 26 msec) SE (TE, 50 msec) SE (TE, 100 msec)

overall

13.5

6.8 ± 2.8

Single-excitation

three

over

without

of pancreas]/standard

ground

pancreatic placed

signal intensity of the of interest ( > 25 pixels) patients with pancreatic to calculate lesion-pancreas noise ratio (C/N) (C/N

rank

one

T2W SE (2,500/100,

Images

of back-

RESULTS

excitation)

by

signal

exact test. Variations in S/N among diffenent imaging methods were evaluated with the nonpanametnic Wilcoxon signed

tations)

Image

loops

= 8) four exci-

T1W SE (250/20,

all

de=

MR (n

Conventional

The

the were

of bowel

intensity

Analysis S/N

Imaging Technique and Sequence

Values

(S/N

noise). Intraluminal bowel signal was measured in small bowel

Visibility

*

lumen,

standard

deviation

gions

out

Table 1 Quantitative of Pancreatic

measure-

on bowel

divided

loops

men

duct

was provided of the pan-

intensity

intensity/standard

-

pancreatic

S/N

of background

ground intensity

qualita-

lumen in the region S/N values were cal-

pancreas

respectively, viation

bowel tail.

signal

of the

In

duct.

evaluation of the

from

ments

the

bile

Quantitative measurement

pan-

of anatomic

including

creas and small of the pancreatic

L.

assessed

identification

common

culated

of the

and tail from bowel.

Image

varying

TE,

or

Analysis differences

sue signal intensity related cay were seen on SE images. barly, variation of TI produced

in tisto T2 deSimiTi-

related effects on IR images. Table 1 summarizes the S/N of the pancreas for the three conventional pulse sequences and the nine echo-planar imaging techniques. The single-excitation SE sequence with a TE of 26 msec

provided

the

highest

(P

< .05)

pancreatic S/N. Within each given set of T2W sequences, increase in TE resulted in decreased S/N (Table 1). The high-resolution single-excitation technique showed significantly higher S/N than did the maximum-resolution two-excitation technique (P < .05). On T1W images, the short TI (100 msec) and long TI (800 msec) IR sequences had a significantly (P < .05)

higher with

than

did

a TI of 380

S/N

msec

the sequence (which

was

Radiolosrv

#{149} 177

chosen to minimize or eliminate signab from pancreas and liver). Table 2 summarizes quantitative measurements of bowel lumen enhancement. Intraluminab signal intensity, which was high with all of the pulse sequences evaluated, was significantly higher (P < .05) after oral administration of the bowel contrast agent. The single-excitation T2W SE technique demonstrated a significantly higher S/N (P < .05) than did the two-excitation T2W SE technique. Fluid-filled bowel appeared hyperintense on T1W IR images because the TI used was short compared with the long Ti of the aqueous intrabuminab bowel content.

Table 2 Bowel Lumen

Enhancement

Imaging Technique

and

Unenhanced

Sequence

S/N

Single-excitation

S/N

(n = 18)*

Enhancementt

TIW

IR (TI, 100 msec) IR (TI, 380 msec) IR (TI, 800 msec) Single-excitation

3.9

± 2.0 2.5 ± 1.2 1.5 ± 0.7

47.3 36.5 23.9

± 7.9

7.0 ± 3.1 4.8 ± 2.1 3.3 ± 1.2

55.1 52.3 50.1

± 7.7

4.8

33.4 25.8 24.9

12.1 14.6 15.9

± 7.5 ± 5.7

T2W

SE (TE, 26 msec) SE (TE, 50 msec) SE (TE, 100 msec) Two-excitation T2W

.

Enhanced

18)

(n =

SE (TE, 26 msec) SE (TE, 50 msec) SE (TE, 100 msec)

± 2.7 2.6 ± 1.0 1.7 ± 0.5

Note-Values are expressed as mean ± 1 standard * SI was increased significantly in the postcontrast t Enhancement is expressed as the factor by which

7.9

11.0

±

10.9

± 10.8

16.5

± 5.7

7.0 9.9 14.7

± 2.6 ± 2.6

deviation. group with all pulse sequences the intraluminal S/N increased.

studied

(P < .05).

DISCUSSION In the

current

study,

of echo-planar

the

normal

Signal

the

potential

MR

imaging

pancreas

was

evaluated.

enhancement

of

intensity

to depict

bowel lumen and different echo-planar imaging techniques were used. Echo-planar images are generally bebieved to be limited by a low S/N (8,11,14). Our study demonstrates, however, that abdominal MR imaging with echo-planar techniques provides excellent S/N, probably because motion noise is suppressed by the short acquisition time. study, pancreas

Therefore, S/N actually

in this ex-

ceeded the S/N of our conventional T2W images (Table 1) or those in the literature (19,20). Echo-planar MR imaging provided pancreatic images free of motion artifacts (Fig 1) and thus provided images of improved diagnostic quality. A section thickness of 10 mm and a gap of 5 mm yielded images with an acceptable trade-off of anatomic coverage and S/N. Singleexcitation T2W images, each obtained in 40 msec, showed the overall best performance

organ

in sharp

delineation

of

margins.

The oral administration

of a high-

bowel contrast agent improved the distinction of pancreas from adjoining bowel (Fig 2) and enhanced intraluminab S/N by an average factor of 10-15 (Table 2). Long T2 aqueous gastrointestinal contrast media may well serve as bowel enhancement agents in echo-planar imaging. Increased signal in the bowel lumen signal-intensity

contributes

image

noise

because

the

image

acquisition

nar

image

acquisition

is extremely

short

bowel (21).

time, compared

moves

dur-

Echo-pla-

however, with

all

gross physiologic motion. High signal intensity in intrabuminal bowel is therefore achieved without cost to 178

#{149} Radioboirv

hancement values are tional MR

tensity

liminary

retropenitoneum

and

high-

is feasible. Pancreatic S/N similar to those in convenimaging. Thus, in this pre-

study,

signal-intensity bowel is qualitatively similar to that on CT scans. Studies to investigate the capability of aqueous bowel contrast agents to enhance the entire bowel are currently under way. Quantitative data demonstrated that the single-excitation T2W echoplanar imaging technique with a TE of 26 msec had a significantly (P < .05) higher pancreatic S/N than achieved with two-excitation T2W or single-excitation T1W echo-planar techniques or conventional SE techniques. Multisection two-excitation techniques require a breath hold of about 12 seconds, making them more susceptible to spatial misregistration. There is, however, a trade-off between anatomic resolution and S/N. When the matrix size is increased from 128 x 128 to 256 x 128, the S/N is expected to decrease as a result of the twofold reduction in voxel vobume. Although the expected decrease in S/N is only 21/2, our experimental data demonstrate somewhat lower

are available to the detection

S/N,

pancreas

presumably

as a consequence

of

abdominal tions and

motion between excitathe decrease in TR (from to 6 seconds). Since high S/N and

C/N

are

both

important

ing, single-excitation preferable for routine pancreas. It appears lect

to a conventional

ing

overall image quality. The contrast pattern on fat-suppressed echo-pbanar MR images with a low-signal-in-

images

at a variety

in MR

techniques imaging necessary of TEs:

of focal

imag-

seem of the to cob-

pancreatic

prerequisites

this technique characterization

lesions.

These

le-

might be detected and identified on the basis of their difference in signab behavior and good tissue contrast (Fig 3), as has previously been demonstrated with hepatic MR imaging sions

(9). IR techniques from

nulbing

normal

the

pancreatic

be advantageous (22), despite their

signal

tissue

for lesion

might

detection

limited S/N, belesions have different null therefore appear hypenintense with this TI (22). A prospective controlled study comparing CT performed with bowel opacification and high-resolution echo-planar MR imaging performed with bowel en-

cause points

focal and

hancement is currently in progress. In summary, our study illustrates that trast

pancreatic imaging material-enhanced

tion

echo-planar

allows

and oc

all of the to apply and

imaging

reliable

of the

surrounding

improve

con-

is feasible,

differentiation

from

may

with oral high-resolu-

bowel,

the diagnosis

of

pancreatic disease. High S/N, elimination of motion artifacts, and intrabuminal bowel signal enhancement are critical

steps

in

developing

strategy

for abdominal

imaging.

#{149}

an

overall

echo-planar

Short

TE images yield higher S/N but limited C/N, whereas long TE images have improved C/N at the cost of neduced S/N. We have demonstrated that echoplanar imaging of the pancreas with adequate bowel lumen signal en-

Acknowledgment: PhD, for his statistical

We thank analysis.

Elkan

Halpern,

References 1.

Chezmar

Hulnik Liver

JL,

DH, and

Rumancik

Nelson abdominal

WM,

Megibow

RC, Bernardino screening

AJ,

ME. in patients

Ianuarv

1992

2.

with cancer: ology 1988;

CT versus 168:43-47.

MR imaging.

Ehman

McNamana

MT.

RL,

9.

Radi-

4.

5.

Henkelman

RM,

Bnasch

RC,

6.

7.

8.

Bronskill

MJ.

Artifacts

10. in

magnetic resonance imaging. Rev Magn Reson Med 1987; 2:1-126. Saini 5, Modic MT. Hamm B, Hahn PF. Advances in contrast-enhanced MR imaging. AJR 1991; 156:235-254. Ungen EC, Cohen MS, Gatenby RA, et al. Single breath-holding scans of the abdo-

men using

FISP and FLASH

MJ,

Gastrointestinal with echo-planar

Evans

DF,

Lamont

12.

13.

14. 15.

#{149} Miimhor

1

nance

imager.

17.

Society

of Magnetic

of Magnetic

Reso-

1987; 51.

19.

ods in medical research. Oxford, Blackwell Scientific, 1987. Tscholakoff D, Hnicak H, Thoeni kler ML, diagnosis

Margulis AR. of pancreatic

England: R, Win-

MR imaging in the disease. AJR 1987;

148:703-709.

20.

21.

22.

Steiner

E, Stark

DD, Hahn

PF, et al.

Imag-

ing of pancreatic neoplasms: comparison of MR and CT. AJR 1989; 151:487-491. Hahn PF, Stank DD, LewisJM, et al. First clinical trial of a new superparamagnetic iron oxide for use as an oral superparamagnetic contrast agent in MR imaging. Radiology 1989; 175:695-700. Bydder GM, Young IR. MR imaging: dinical use of the inversion recovery sequence. J Comput Assist Tomogn 1985; 9:659-675.

in Medi-

cine, 1989; 836. Rzedzian RR. High speed, high tion, spin echo imaging by mosaic

In: Book of abstracts: Resonance

Society

18.

RM.

Resonance

Calif:

in Medicine,

Rzedzian RR. Method of high speed imaging with improved spatial resolution using partial k-space acquisitions. U.S. patent 4,767,991, 1988. Armitage P. Berry C, eds. Statistical meth-

Radiology

Fat suppression techniques: a comparison of results in instant imaging (abstr). In: Book of abstracts: Society of Magnetic Resonance in Medicine 1989. Berkeley, Calif:

16.

Berkeley,

A, Atkinson

TunboFLASH:

Ultra-fast imaging. Magn Reson Imaging 1991; 9:1-37. Weisskoff RM, Cohen MS. Rzedzian RR.

of Magnetic

1R9

pediatric

1988; 166:157-163. Cohen MS. Weisskoff

mesh (abstr).

Vnliimo

B, Singer

Segmented

Method for breath-hold MR imaging of the liver with flexible contrast. Radiology 1990; 177:515-521. Rzedzian RR, Mansfield P. Doyle M, et aI. Real-time NMR clinical imaging in pediatrics. Lancet 1983; 2:1281-1282. Saini 5, Stank DD, Rzedzian RR, et aI. Forty-millisecond MR imaging of the abdomen at 2.0 T. Radiology 1989; 173:111-116. Stehling MJ, Howseman AM, Ordidge RJ, et al. Whole-body echo-planar MR imaging at 0.5 T. Radiology 1989; 170:257-263. Crooks LE, Arakawa M, Hylton NM, et al.

Echo-planar

G, et al.

tract: dynamic MR studies imaging. Radiology 1989;

171:41-46.

ii.

at 1.5 Tesla.

Comput Tomogr 1988; 12:575-583. Rzedzian RR, Pykett IL. Instant images of the human heart using a new, whole body MR imaging system. AIR 1987; 149:245-250. Pykett IL, Rzedzian RR. Instant images of the body by magnetic resonance. Magn Reson Med 1987; 5:563-571. Stehling

RR, Wailnen

Di, Saini S.

Felmlee JP, Gray JE, Higgins CB. Influence of physiologic motion on the appearance of MR images. Radiology 1986; 159: 777-782. 3.

Edelman

in Medicine

nesoluscan and

Society 1987.

.

-

#{149}

70