Comparisons of water-equivalent diameter measured on

Comparisons of water-equivalent diameter measured on

1

images of abdominal routine computed tomography

2

with and without a contrast agent

3

Arrum Nitasari¹, Choirul Anam^1*, Wahyu Setia Budi¹, Annisa Lidia Wati¹, Syarifudin

4

Syarifudin², Geoff Dougherty³

5

1Department of Physics, Faculty of Mathematics and Natural Sciences, Diponegoro University, Jl. Prof. Soedarto SH, Tembalang,

6

Semarang 50275, Central Java, Indonesia

7

2Department of Radiology, Dr. Kariadi Hospital, Jl. Dr. Sutomo No.19, Semarang Selatan, Semarang 50244, Central Java,

8

Indonesia

9

3Department of Applied Physics and Medical Imaging, California State University Channel Islands, Camarillo, CA, USA

10 11

A R T I C L E I N F O A B S T R A C T

12 13

AIJ use only:

Received date Revised date Accepted date

Keywords:

Abdominal routine

Size-specific dose estimate (SSDE) Water-equivalent diameter (DW)

The size-specific dose estimate (SSDE) is an estimation of patient dose in computed tomography (CT). The SSDE strongly depends on the water-equivalent diameter (DW). In abdominal CT examinations, a contrast agent is sometimes used to more clearly visualize tissue lesions. The Hounsfield unit (HU) of CT images with and without the use of a contrast agent at specific areas is different and it may affect the DW value. This study aims to compare the DW values calculated from axial CT images in patients who had undergone routine abdominal scans both with and without the use of a contrast agent. Axial images of 144 patients with a weight range of 3.5 kg to 90 kg who had undergone routine abdominal scans both with and without the use of a contrast agent using a Siemens Sensation 64 CT scanner were retrospectively collected. The DW values were automatically calculated using the Matlab-based IndoseCT (version 15a) software. The results showed the percentage difference between DW,contrast and DW,non-contrast is below 2%. As a result, the mean SSDEcontrast is 1.5% smaller than SSDEnon-contrast. Due to the effect of a contrast agent on the DW and SSDE values is below 2%, the axial images of CT abdomen without the use of a contrast agent can be used as the accurate estimation of DW and SSDE for images with the use of a contrast agent.

© 2020 Atom Indonesia. All rights reserved

14 15

INTRODUCTION^

16

17

Computed Tomography (CT) was introduced

18

in the 1970s, and has had a crucial role in the

19

diagnostic field due to its excellent image quality [1,

20

2]. Image acquisition in diagnostic CT is very fast

21

due to the advances in acquisition techniques such as

22

helical and multi-slice CT (MSCT). These

23

advantages lead to the growing use of CT innumerous

24

applications. However its disadvantage is that it

25

contributes to the highest radiation exposure in the

26

medical field [3–5]. Hence, its implementation must

27

be carefully and prudently optimized so that its

28

advantages outweigh this disadvantage.

29

Estimating accurate patient dose is important.

30

Accurate estimation of patient dose relies on the

31

Corresponding author.

metric known as the size-specific dose estimate

32

(SSDE) [4–6]. Since SSDE was announced, several

33

softwares have been developed to measure the

34

effective dose and organ dosage based on SSDE [9,

35

10]. Several studies reported that SSDE was used to

36

represent dose optimization values such as diagnostic

37

reference level (DRL) and acceptable quality dose

38

(AQD) [11–14]. SSDE is estimated from the output

39

dose or the volume CT dose index (CTDIvol) and

40

patient characteristics. In this case, the patient

41

characteristics include X-ray attenuation and patient

42

size. The X-ray attenuation is based many factors,

43

such as type of material and its density and energy of

44

X-ray beam. The X-ray attenuation is fundamental

45

parameter effecting X-ray absorption and dose

46

absorbed by patient. The patient characteristics are

47

E-mail address: anam@fisika.fsm.undip.ac.id DOI: provided by AIJ

generally represented by the water-equivalent

48

diameter (DW) [15, 16] of the patient. Its value can be

49

determined from patient images [7,17,18].

50

X-ray attenuation in CT is expressed in

51

Hounsfield Units (HU). Each body tissue has

52

different HU values, however the resulting contrast

53

for distinguishing adjacent tissue may be relatively

54

small. Thus, a contrast agent is needed to improve the

55

image contrast of the desired body tissue [19, 20]. In

56

a CT examination, the contrast agent is iodine-based

57

[21, 22]. Common indications for the use of a

58

contrast agents on CT abdominal scans include acute

59

appendicitis, cancer staging, diverticulitis, and

60

pancreatitis [23].

61

Because the pixel values (HU) within image

62

change in some tissues before and after agent contrast

63

injection, the DW and the SSDE may change as a

64

result. Based on our best knowledge, no previous

65

study investigated the Dw and SSDE before and after

66

contrast agent injection. As consequence, the Dw and

67

SSDE before and after contrast agent injection are

68

considered the same without justification from the

69

careful study. Therefore, this study aims to calculate

70

the DW values based on axial CT images of a CT

71

examinations with and without the use of a contrast

72

agent. In addition, we also evaluated the SSDE value

73

before and after contrast agent injection. The results

74

of this study will be useful to confidently consider the

75

Dw and SSDE for CT examinations with a contrast

76

agent administrations.

77 78 79

MATERIALS AND METHODS

80

81

Patient images

82

We calculated the DW values of 144 patients

83

with a weight range of 3.5 kg to 90 kg who underwent

84

routine abdominal CTs both with the use of a contrast

85

agent referring to the portal venous phase and without

86

the use of a contrast agent. The contrast agent was

87

Iopamidol 370 (IOP 370). The patient was scanned

88

using the Siemens Sensation 64 CT scanner installed

89

at Dr. Kariadi Hospital, Semarang, Central Java,

90

Indonesia. The routine abdominal protocol for

91

pediatric patients, with and without the use of a

92

contrast agent, was 120 kVp, 85 mA, 1.4 pitch factor,

93

and 28.8 mm total collimation width. For adult

94

patients the current was increased to 200 mA current,

95

with all other settings the same. All the exposure

96

parameters mentioned were obtained from each

97

patient's DICOM header. In addition, both the

98

abdominal CT examination protocol in children and

99

adults applied the tube current modulation (TCM)

100

technique. The TCM was to manage tube loading to

101

optimize dose delivery. Once the TCM was used, the

102

tube loading for each slide would be different due to

103

different patient equivalent thickness. While the

104

CTDIvol were obtained from the CT dose summary as

105

shown in Fig. 1.

106 107

Calculation of DW

108

The DW of a patient can be calculated from

109

the mean HU in the region of interest (ROI). DW was

110

calculated using Eq. (1).

111



HU A

D_w 

 



1

1000

2 1

(1)

112

where 𝐴 is the area of the patient and 𝐻𝑈̅̅̅̅ is the mean

113

HU value within the patient [24]. In this study, we

114

used axial image data extracted from each patient’s

115

dicom files to calculate the DW value. The DW was

116

automatically calculated using Matlab-based

117

IndoseCT version 15a software [18, 25]. The

118

IndoseCT will automatically make the contouring as

119

shown in Fig. 2 and calculate the DW according to the

120

selected tool. For the DW calculation on the IndoseCT

121

we used the “3D” tool and selected a slice number of

122

9 (in Fig. 3), i.e. 9 slices were selected from the axial

123

CT images to represent the water-equivalent diameter

124

along the longitudinal axis. A previous study [6]

125

found that DW values calculated using these settings

126

had a mean percentage difference value less than 1%

127

compared to the DW from all slices[6]. DW value

128

calculations for each patient were performed using

129

axial CT images with and without the use of a

130

contrast agent.

131 132

Calculation of SSDE

133

SSDE values can be calculated by CTDIvol

134

which can be obtained from a CT scanner dose

135

summary [18] and the size-conversion factor (f).

136

f CTDI

SSDE _vol (2)

137

The size-conversion factor of DW can be obtained

138

from the AAPM report 220 [7]. In this study, both

139

SSDE values (SSDEcontrast and SSDEnon-contrast) were

140

automatically calculated using the IndoseCT version

141

15a [18].

142

143

144

Fig. 1. Captured CT dose summary from Siemens Sensation 64 CT scanner

145 146

147

148

Fig. 2. The auto-contouring by the IndoseCT. (a) original image and (b) auto-contouring result.

149 150

151

Fig. 3. The Dw calculation using 3D tool and a slice number of 9 using the IndoseCT.

152

153 154

Relationship between DW,contrast and DW,non-contrast

155

Comparisons of DW,contrast to DW,non-contrast

156

The number of the DW values calculated

157

using axial CT images with the use of a contrast agent

158

(DW,contrast) referring to the portal venous phase and

159

without the use of a contrast agent (DW,non-contrast) was

160

144 respectively. The DW,contrast and the DW,non-contrast

161

values for abdominal scans were compared. The

162

percentage difference between DW,contrast and DW,non-

163

contrast was calculated using Eq. (3).

164

100

, ,

, 









 





 contrast non W

contrast non W contrast W

D D

PD D (3)

165 166 167

RESULTS AND DISCUSSION

168

169

Fig. 4 presents the axial CT images of the

170

abdomen with and without a contrast agents. It shows

171

that the axial CT images of the abdomen with the use

172

of a contrast agent provides better visualization of the

173

portal vein and solid organs, i.e. the liver, spleen,

174

kidney, and pancreas than the axial CT images of the

175

abdomen without the use of a contrast agent due to

176

the HU values increase in these areas [27, 28]. The

177

mean HU value in some areas of the axial CT image

178

of the abdomen was 22.6% greater than without the

179

use of a contrast agents.

180 181 182

Relationship between DW,contrast and DW,non-contrast

183

Comparisons of DW,contrast to DW,non-contrast

184

The relationship between DW,contrast and

185

DW,non-contrast is shown in Fig. 5a. The correlation is

186

strong (R² = 0.974) which indicates that two Dws have

187

strong correlations. Based on this finding, the

188

DW,contrast can be accurately calculated from DW,non-

189

contrast. Fig. 5b shows the box-plot diagrams of

190

DW,contrast and DW,non-contrast. The mean value of

191

DW,contrast is 1.2% greater than the mean value of

192

DW,non-contrast. This indicates that there are only small

193

effect of a contrast agent on the DW value because the

194

use of a contrast agent leads to an increase the HU

195

values only in some organs [19, 20, 26]. This finding

196

is consistent with the very recently reported study

197

[29]. They reported that Dw after contrast agent

198

injection is only 0.21 cm greater that those before

199

contrast agent injection.

200 201 202

203

204

Fig. 4. (a) The axial CT image of the abdomen without the use of a contrast agents, and (b) The axial CT image of the

205

abdomen with the use of a contrast agents showing the liver; PV, portal vein; IVC, Inferior vena cava; A, aorta; S,

206

stomach; P, pancreas; K, kidney, and spleen.

207

208

209

210

211

212

213

214

215

216

217

218

219

220

8 12 16 20 24 28 32 36

8 12 16 20 24 28 32 36

DW,contrast

DW,non-contrast

974 .

2 0 R

308 . 0 998

.

0 _,

,_contrast  _{wnoncontrast}

w D

D

(a)

DW,non-contrast D_W,contrast

10 15 20 25 30 35

MIN= 10.4 MEAN= 23.4 MAX= 33.4

MIN= 10.5 MEAN= 23.2

DW (cm)

(b)

MAX= 32.8

Fig. 5. (a) Relationship between DW,contrast and DW,non-contrast;and (b) Box-plots of DW,contrast and DW,non-contrast.

221 222 223

Relationship between CTDIvol,contrast and

224

CTDIvol,non-contrast

225

The relationship between CTDIvol,contrast and

226

CTDIvol,non-contrast is shown in Eq. 6a. The correlation

227

is very strong (R² = 0.995) which indicates the values

228

of CTDIvol,contrast and CTDIvol,non-contrast have strong

229

correlation. Box-plots in Fig. 6b show a very small

230

difference (0.1%), with the mean value of

231

CTDIvol,contrast greater than the mean value of

232

CTDIvol,non-contrast. This is not significant since the

233

same protocol scan is used for routine abdominal

234

scans with and without the use of a contrast agent.

235

For information, the CT scanner used at our

236

institution applies the TCM technique. In the TCM

237

technique, the CTDIvol is affected by the tube loading

238

and the tube loading is affected by the size of the

239

patient. The correlation between tube loading and

240

CTDIvol had been reported by the previous study [18].

241

Relationship between SSDEcontrast and SSDEnon-

242

contrast

243

The relationship between SSDEcontrast and

244

SSDEnon-contrast is shown in Fig. 8a. The correlation is

245

strong (R² = 0.943). The box-plots (Fig. 8b) show that

246

the mean value of SSDEcontrast is 1.5% smaller than

247

the mean value of SSDEnon-contrast, which is due to

248

DW,contrast being 1.2% greater than the mean value of

249

DW,non-contrast. Therefore, the different of SSDE before

250

and after contrast agent is not significant. This result

251

is consistent with the very recent study reported that

252

SSDE from images with contrast agent is rougly 1%

253

lower than those from images without contrast agent

254

[29]. AAPM reported that if difference between the

255

pre- and post-scan values of DW and SSDE is below

256

10%, the final values used may be taken from the pre-

257

scan values [7].

258 259 260 261

2 4 6 8 10 12 14

2 4 6 8 10 12 14

CTDIvol,contrast (mGy)

CTDIvol,non-contrast (mGy) 995

.

20 R

037 . 0 995

.

0 _,

,_contrast  _{volnoncontrast}

vol CTDI

CTDI

(a)

SSDEnon-contrast SSDE_contrast

6 8 10 12 14 16 18 20 22 24

MIN= 7.9 MEAN= 13.8 MAX= 21.6

MIN= 7.8 MEAN= 14.0 MAX= 21.8

SSDE (mGy)

(b)

Fig. 6. (a) Relationship between CTDIvol,contrast and CTDIvol,non-contrast, and (b) Box-plots for CTDIvol,contrast and

262

CTDIvol,non-contrast.

263

264 265

266

6 8 10 12 14 16 18 20 22 24

6 8 10 12 14 16 18 20 22 24

SSDEcontrast (mGy)

SSDEnon-contrast (mGy) 943

.

2 0 R

782 . 0 931

.

0  

 _{noncontrast}

contrast SSDE

SSDE

(a)

CTDIvol,non-contrast CTDIvol,contrast 2

4 6 8 10 12 14

MIN= 3.4 MEAN= 9.11 MAX= 13.6 MAX= 13.6

MEAN= 9.12

MIN= 3.4

CTDIvol (mGy)

(b)

Fig. 7. (a) Relationship between SSDEcontrast and SSDEnon-contrast, and (b) Box-plots for SSDEcontrast and SSDEnon-contrast.

267 268 269

CONCLUSION

270

271

As expected, DW,contrast is greater than DW,non-

272

contrast. The percentage difference between DW,contrast

273

and DW,non-contrast is 1.2%. As a result, the mean

274

SSDEcontrast is 1.5% smaller than SSDEnon-contrast. Due

275

to the effect of a contrast agent injection on the DW

276

and SSDE values is not significant (below 10%), the

277

axial images of CT abdomen without the use of a

278

contrast agent can be used to estiamate the DW and

279

SSDE from images with the use of a contrast agent

280

injection.

281 282 283

ACKNOWLEDGEMENT

284

285

This work was funded by the World Class

286

Research University (WCRU), Diponegoro

287

University, 2021.

288 289 290

AUTHOR CONTRIBUTION

291

292

A.N. First_author and C.A. Second_author

293

equally contributed as the main contributors of this

294

paper. All authors read and approved the final version

295

of the paper.

296 297 298

REFERENCES

299

300

[1] G.D. Rubin, Radiology 273 (2014) S45–S74.

301

[2] UNSCEAR, I (2010) 156.

302

[3] G. Kirova and G. Hadjidekov, Rentgenol. i

303

Radiol. 44 (2005) 166–175.

304

[4] S.P. Power, F. Moloney, M. Twomey et.al.,

305

World J. Radiol. 8 (2016) 902.

306

[5] United Nations Scientific Committee on the

307

Effects of Atomic Radiation, Sources and

308

Effect of Ionizing Radiation, United Nations,

309

New York, 2010.

310

[6] C. Anam, F. Haryanto, R. Widita et.al., J.

311

Appl. Clin. Med. Phys. 17 (2016) 320–333.

312

[7] AAPM, AAPM Rep. 220 (2014) 1–23.

313

[8] R.D.A. Khawaja, S. Singh, B. Vettiyl et.al.,

314

Am. J. Roentgenol. 204 (2015) 167–176.

315

[9] P. Sahbaee, W.P Segars and E. Samei, Med.

316

Phys. 41 (2014) 072104.

317

[10] A. Fahmi, C. Anam, Suryono et.al., Radiat.

318

Prot. Dosimetry 188 (2020) 522-528.

319

[11] R. Imai, O. Miyazaki, T. Horiuchi et.al.,

320

Pediatr. Radiol. 45 (2015) 345–353.

321

[12] A. Mohammadbeigi, K. Khoshgard, A.

322

Haghparast et.al., Radiat. Prot. Dosimetry

323

186 (2019) 496–506.

324

[13] A. Mehdipour and M. Parsi, Radiat. Prot.

325

Dosimetry 185 (2019) 176–182.

326

[14] K.M. Kanal, P.F. Butler, D. Sengupta et.al.,

327

Radiology 284 (2017) 120–133.

328

[15] M. Bostani, K. McMillan, P. Lu el.al., Med.

329

Phys. 42 (2015) 958–968.

330

[16] C. Anam, D. Adhianto, H. Sutanto et.al., J.

331

Xray. Sci. Technol. 28 (2020) 695–708.

332

[17] C. Anam, F. Haryanto, R. Widita et.al., J.

333

Phys. Conf. Ser. 1204 (2019) 012002.

334

[18] C. Anam, F. Haryanto, R. Widita et.al., Int. J.

335

Radiat. Res. 16 (2018) 289–297.

336

[19] E. Amato, I. Salamone, S. Naso et.al., Am. J.

337

Roentgenol. 200 (2013) 1288–1293.

338

[20] L. Hrvoje and M.W. Greenstaff, X-Ray

339

Computed Tomography Contrast Agents,

340

2014.

341

[21] P. Sahbaee, E. Abadi, W.P. Segars et.al.,

342

Radiology 283 (2017) 749–757.

343

[22] A. Suwannasri, R. Kaewlai and S.

344

Asavaphatiboon, J. Phys. Conf. Ser. 694

345

(2016).

346

[23] J. V. Rawson and A.L. Pelletier, Am. Fam.

347

Physician 88 (2013) 312–316.

348

[24] C. Anam, I. Arif, F. Haryanto et.al., Radiat.

349

Prot. Dosimetry 185 (2019) 42–49.

350

[25] C. Anam, F. Haryanto, R. Widita et.al., J.

351

Phys. Conf. Ser. 694 (2016) 012030.

352

[26] K.T. Bae, Radiology 256 (2010) 32–61.

353

[27] K. Guite, Louis and F. Lee, Sel. Top. Comput.

354

Tomogr. (2013) 1–14.

355

[28] K. Ishigami, T. Tajima, A. Nishie et.al., Eur.

356

J. Radiol. 74 (2010) e95–e100.

357

[29] B Viggiano, S. Rose and T.P. Szczykutowicz,

358

Med. Phys. (2021). doi: 10.1002/mp.14721

359

360 361

362