Stomach and Duodenum: Radiographic Magnification Using … · 2015. 7. 29. · c. a. b. c. Volume 160 Number 2 Radiology #{149}385 Figure 5. (a)The compression technique reveals asmall

Yoshihisa Nakano, MD #{149}Kaori Togashi, MD #{149}Kazumasa Nishimura, MDa Kyo Itoh, MD #{149}Ichiro Fujisawa, MD #{149}Reinin Asato, MD #{149}Hideki Adachi, MD

#{149}Harumi Itoh, MD #{149}Kanji Torizuka, MD

Stomach and Duodenum:Radiographic Magnification Using

Computed Radiography (CR)’

383

Abdominal and Gastrointestinal Radiology

We performed direct radiographicmagnification (X3) using a 0.1-mmmicrofocal tube and computed radi-ography (CR) in air-barium double-contrast studies of the stomach andduodenum. To eliminate blurring ofthe image due to motion, we usedthe maximum kilovolt peak ratepossible (102 kVp), the maximummilliampere second rate possible (64mAs), the shortest possible exposuretime (0.1 second), and a horizontalx-ray beam. With the patients in thesupine position, CR provided thewide image reproduction range re-quired to obtain satisfactory radio-graphs. Image processing in the CRsystem produced radiographs withincreased radiographic contrast andenhanced edge definition of ana-tomic borders or pathologic process-es. The duodenal villi were clearlyvisible in 45% of the cases.

Index terms: Castrointestina! tract, radiogra-

phy, 7.12 #{149}Magnification #{149}Radiography, digi-

ta!

Radiology 1986; 160:383-387

I From the Department of Radiology and Nu-clear Medicine, Kyoto University Medical

School, Sakyo-ku, Kyoto 606, Japan. ReceivedDecember 13, 1985; revision requested Febru-ary 3, 1986; revision received February 24; ac-cepted March 5. Address reprint requests toY.N.

C RSNA, 1986

T HE potential of radiographic mag-

nification to facilitate diagnosis

has been recognized for many years.

Direct radiographic magnification

has already been used to help diag-

nose diseases of the breast and skele-

ta! soft tissue. However, theme are in-

herent differences between obtain-

ing radiographic magnification im-

ages of the gastrointestinal (GI) tract

and obtaining images of other body

systems (1). Adequate radiation pene-

tration of the thickest part of the

body requires the use of a limited x-

ray tube current and small focal-spot

size, and a short exposure time must

also be employed to avoid motion am-

tifact, a major problem in imaging

the GI tract. In addition, a higher ki-

lovolt peak (kVp) is required so that

the column of barium-sulfate suspen-

sion can be penetrated by the x rays.

The techniques that have been suc-

cessfully used in skeletal and mam-

mographic magnification radiogra-

phy must be modified to accom-

modate these differences.

We had the patients in our study

assume the supine position, and we

used a horizontal x-ray beam to ob-

tam the magnification radiographs.

With the patients in the supine posi-

tion, we were able to visualize the

stomach and the duodenum without

their overlapping the barium-filled

GI tract, the lumbar spine, and the

thick muscles of the back (Fig. 1). Ad-

equate radiation penetration was ob-

tamed despite the system’s low gen-

erator capacity, which was due to

tube limitations.

Although the horizontal view

makes direct radiographic magnifica-

tion possible, the detector systemmust have a wide latitude since the x

ray has to penetrate the thick part of

the upper abdomen and the air-filled

stomach at the same time. Conse-

quently, it is technically difficult to

obtain optimum density using a con-

ventional screen-film system . We

have used the Fuji (Kanagawa, Japan)

computed radiography (CR) system

to obtain radiographic magnification

images of the GI tract and have

found it useful in helping diagnose

disorders in a select group of patients

in whom equivocal findings were ob-

tamed in standard upper GI tract ex-

aminations (2).

MATERIALS AND METHODS

We used an angiographic system con-sisting of a Siemens (Erlangen, West Ger-

many) Tridoros i000 generator and a C-

arm arrangement that employed a

microfocal tube (OPTI 1 iO/ 12/50/HSG)with a nominal focal spot of 0.1 mm andan image intensifier. When the CR’s C-

arm arrangement is used, the cassette

with the imaging plate is attached to the

front of the image intensifier (Fig. 2).

After the routine upper GI examinationon the tilt-table radiography-flounoscopy

table was completed, each patient in our

study was brought to the angiography ta-ble. The patient assumed the supine posi-

tion (Fig. 2), and horizontally exposed

magnification radiographs using the CR

system were obtained at 102 kVp and 64

mAs, and with an exposure time of 0.i

second. The image was magnified 2.5-3

times the original magnification, with a

focus-film distance of 115 cm and a focus-

object distance of 40 cm.

In the past 2 years, we have performed

upper GI tract examinations on 520 pa-

tients, and we have fluoroscopically and

radiographically identified equivocal

findings in 95 patients. To establish more

conclusively the pathologic diagnosis of

these 95 patients, we obtained an addi-

tional magnification radiograph using ahorizontal x-ray beam. The decision to

obtain an additional projection was usual-

ly made during the fluoroscopic examina-

tion.

The demonstrability of the villi on the

duodenal bulb was also evaluated in 80

patients in whom no duodenal abnormal-

ities were found, since the demonstration

of micromucosal patterns, such as areae

gastnicae or duodenal villi, is a bench-

mark by which our technique can be eval-

uated (3).

1. 2.

384 . Radiology August

Figures 1, 2. (1) Horizontal view of the stomach and duodenum with patient in supine position. (2) Patient in supine position on the an-

giography table. When the C-arm arrangement is used, the cassette with the imaging plate is attached to the front of the image intensifier

(arrow).

a. b.

Figure 3. (a) Magnification radiograph (X3) obtained with the CR system. (b) Close-up

view shows the fine network pattern of areae gastnicae.

Figure 4. Micromucosal pattern of duode-

nal villi.

RESULTS

Sixty-eight of the 95 magnification

radiographs obtained using the CR

system were of excellent quality, and

visualization of the mucosal pattern

was superior to that demonstrated by

the nadiographs obtained during theroutine examinations (Fig. 3). Twen-

ty-two images were acceptable for di-

agnostic use, although use was limit-

ed by mild image noise (ten studies),

slight motion artifact (eight studies),

on improper patient positioning (four

studies). Five studies were not ac-

ceptable due to movement on the

pant of the patient on technical error

(i.e., misselection of data acquisition

parameters).

We determined that duodenal villi

were demonstrated in images ob-

tamed from 36 of 80 patients (45%).

pattern that resembled a shaggy car-

pet was seen on the magnificationdiographs; that is, multiple finger-

like projections seen en face in the

duodenal bulb (Fig. 4) (1).

In 70 of the 90 patients from

c.

a. b. c.

Volume 160 Number 2 Radiology #{149}385

Figure 5. (a) The compression technique reveals a small polyp of the stomach (arrow). (b) Double-contrast study does not reveal any lesions

on the posterior wall of the stomach, so that the lesion appears to be located on the anterior wall. (c) Horizontal magnification using CR re-

veals two polyps (arrows) located on the anterior wall of the antrum of the stomach.

a. b.Figure 6. Early carcinoma on the anterior wall of the stomach. (a) Double-contrast study does not reveal the lesion. (b) The compressionmethod reveals the lesion. (c) The horizontal-magnification radiograph provides radiographic visualization of the small lesion of the anteni-or wall (arrow).

acceptable horizontal-magnification

radiographs were obtained, the pres-

ence of lesions was convincingly ex-

cluded. The presence of lesions dem-

onstrated by the CR system in the

remaining 20 patients was subse-

quently confirmed either surgically

or pathologically. The clinical and

pathologic diagnoses of all 90 pa-

tients are shown in Table 1. Horizon-

tal-magnification nadiographs ob-

tamed using the CR system proved

helpful in evaluating lesions that

were not located in the posterior wall

of the stomach (Figs. 5, 6). In one

case, the horizontal-magnification ma-

diographs also provided additional

information concerning the exact ex-

tension of stomach carcinoma

(Fig. 7).

DISCUSSION

The direct-magnification radio-

graphs (X2.5-3) obtained using the

CR system demonstrated fine detailin the small intestinal mucosa. This

detail could not be seen on the radio-

graphs obtained using the conven-

tional screen-film system. The visual-

ization of a fine, lacy, mucosal

pattern in the duodenal bulb mdi-

cates that the vilius pattern has been

successfully imaged (Fig. 4). In our

study, the villus pattern was demon-

strated in 45% of 80 images, which is

an extremely high success rate com-

pared with the 5%-iO% incidencerate reported by Glick et al. (4).

The advantages of the CR system

compared with the conventional

screen-film system are two-fold.

First, the detector of the CR system

has a wide dynamic range of

1:10,000, and a constant optimum

density can be attained whenever

any quantity of x-ray exposures are

being made. This facility to optimize

the exposure of the CR system is im-

pontant in magnification madiogra-

phy, since magnification radiogra-

phy is a technically difficult pro-

cedure. Second, the CR system allows

the images to be enhanced. Imageprocessing in the CR system incompo-

mates high-pass spatial filtering that

results in enhanced edge definition

and increased visibility of the micro-

mucosai pattern.Air-barium double-contrast radiog-

raphy of the stomach and duodenum

has gained acceptance as a routine

method of examination. This tech-nique emphasizes the en face appear-

ance of the CI mucosa for the detec-

tion of superficial abnormalities.However, the technique frequently

fails to image minute mucosal lesions

that are located in the anterior wall

or in the lesser or greater curvature

of the stomach, and that yield normalresults in the upper GI examination.

The horizontal view not only makes

direct magnification possible but also

reveals abnormalities that were not

visible in images obtained during

routine examination (Figs. 5-7).a. b. The disadvantages of CR magnifi-Figure 7. (a) Routine examination of the stomach reveals the carcinoma in the angulus of cation of the GI tract include the ex-the stomach. (b) Horizontal magnification obtained with CR shows the exact extension of pense of the CR system and of the an-the carcinoma toward the anterior wall of the stomach and along the lesser curvature (ar-

rows).

c.

;;gure 8. (a) Low spatial filtering, (b) intermediate spatial filtering, and (c) high-pass spatial filtering.

August 1986386 #{149}Radiology

a. b. c.

Volume 160 Number 2 Radiology #{149}387

Figure 9. (a) Minimum, (b) moderate, and (c) marked enhancement with intermediate spatial filtering. Marked enhanced image shows theprominent areae gastricae, suggesting chronic gastnitis.

giognaphic equipment using a small

focus tube; the limited output of the

CR system; and most important, the

problems inherent in image en-

� hancement. In Figure 8, three images

representing three kinds of spatial

filtering can be seen: low, intermedi-

ate, and high pass. Intermediate fil-

tening yields the best representation

of the areae gastnicae. In Figure 9, the

three kinds of enhancement obtained

with intermediate spatial filtering

are shown: minimum, moderate, andmarked degree of enhancement. A

moderate degree of enhancement

seems to yield the best nepresenta-

tion of areae gastnicae. Thus, image

enhancement has both advantages

and limitations. Proper image en-

hancement provides increased visu-

alization of the anatomic and patho-

logic structures, but too much

enhancement increases image noise,

resulting in a loss of diagnostic infon-

mation or even creating an artifact

that resembles a pathologic process

(Fig. 9).

In summary, we studied the effec-

tiveness of obtaining horizontal-

magnification radiographs of the

stomach and duodenum using a mi-

crofocal x-ray tube and the CR system

and having the patient assume the

supine position. Our success in visu-

alizing micromucosal patterns mayindicate that this radiographic tech-

nique could facilitate diagnosis of GI

tract disease by providing informa-

tion on mucosal structural abnormal-

ities that were not previously detect-

able by more routine methods. U

References

1. Goldberg HI, Gould R, Rosenquist J, Royal

5, Owens R, Silverman S. In vivo demon-stration of small intestinal villi in dogs andmonkeys using radiographic magnifica-tion. Radiology 1982; 142:53-58.

2. Sonoda M, Takano M, Miyahara J, Kato H.

Computed radiography utilizing scanninglaser stimulated luminescence. Radiology1983; 148:833-838.

3. Seaman WB. The areae gastricae. AJR1978; 131:554.

4. Click SN, Cohel VK, Laufer I. Mucosalsurface patterns of the duodenal bulb. Ra-diology 1984; 150:317-322.