This copy is for personal use only. This copy is for personal use only.

ORIGINAL RESEARCH • MEDICAL PHYSICS

To order printed copies, contact reprints@rsna.org To order printed copies, contact reprints@rsna.org

Procedure-Specific CT Dose and Utilization Factors

for CT-guided Interventional Procedures
Kai Yang, PhD • Suvranu Ganguli, MD • Matthew C. DeLorenzo, MS • Hui Zheng, PhD • Xinhua Li, PhD • Bob
Liu, PhD
From the Division of Diagnostic Imaging Physics, Department of Radiology (K.Y., M.C.D., X.L., B.L.), and Center for Image Guided Cancer Therapy, Department of
Interventional Radiology (S.G.), Massachusetts General Hospital, 55 Fruit St, Boston, MA 02114; and MGH Biostatistics Center, Boston, Mass (H.Z.). Received De-
cember 18, 2017; revision requested February 26, 2018; revision received May 1; accepted May 3. Address correspondence to K.Y. (e-mail: kyang11@mgh.harvard.edu).
Conflicts of interest are listed at the end of this article.
See also the editorial by Leng in this issue.

Radiology 2018; 00:1–8 • https://doi.org/10.1148/radiol.2018172945 • Content code:

Purpose: To present procedure-specific radiation dose metric distributions and define quantitative CT utilization factors for CT-
guided interventional procedures.

Materials and Methods: This single-center, retrospective study collected dictation reports and radiation dose data from 9143 consecu-
tive CT-guided interventional procedures in adult patients from 2012 to 2017. Procedures were sorted into four major interven-
tional categories: ablation, aspiration, biopsy, and drainage, each of which was further divided into subcategories. After exclusion, a
total of 8213 procedures (4391 in men and 3822 in women) were divided into 21 subcategories. The mean patient age at examina-
tion for men was 62 years 6 15 (standard deviation; age range, 19–114 years), and for women it was 61 years 6 15 (age range,
19–113 years). Distributions of dose metrics and CT usage–related parameters were analyzed by category with descriptive statistic
outcomes. Quantitative CT utilization factors (which measure average CT usage) for each interventional subcategory were derived
by using total scan length, acquisition count, and number of images.

Results: Interventional CT scans have distinctly different dose metric characteristics from diagnostic CT scans. Wide variations
of dose metrics were observed among subcategories, even within the same major category. For the most frequently performed CT-
guided interventional procedures within each major category, liver ablation, chest aspiration, liver biopsy, and single abdominal
drainage, the median dose-length product was 2351, 657, 1175, and 1125 mGy ∙ cm, respectively. Procedure-specific CT utiliza-
tion factors ranged between 0.6 and 3.6.

Conclusion: This study provides procedure-specific CT dose metric distributions and quantitative CT utilization factors on the basis
of a large number of procedures and categorization of CT-guided interventional procedures.
© RSNA, 2018

Cventional radiology procedures (1–12) including biop-

T is routinely used as a guidance tool for many inter- literature (25,26) as sufficient to provide adequate radia-
tion dose data for a single facility varies from 10 to more
sies, aspirations, drain placements, and thermal ablations, than 50 procedures.
in addition to serving as a diagnostic imaging modality. A logical first step to establish reference dose levels for
Compared with fluoroscopy-guided interventional proce- CT-guided interventional procedures is to collect a large
dures, the advantage of CT-guided interventional proce- number of individual procedures for each procedure type
dures is the cross-sectional visualization and localization and analyze the dose metric distributions. Leng et al (27)
of the targeted anatomy and interventional equipment. reported a study from 571 CT-guided interventional pro-
For patient protection and practice quality improvement, cedures performed at Mayo Clinic. Five major categories
the radiation dose associated with each CT-guided in- of procedures were analyzed: cryoablation, aspiration,
terventional procedure should be closely monitored by biopsy, drainage, and injection. By considering both in-
tracking key parameters including volume CT dose in- termittent and helical acquisitions, patient entrance skin
dex (CTDIvol), size-specific dose estimate (SSDE), scan dose and effective dose were also calculated. By following
length, and dose-length product (DLP). For diagnostic a similar approach, CT dose data from 9143 CT-guided
CT, radiation dose monitoring programs have become interventional procedures consecutively performed at our
mandatory and national reference dose levels are being institute within last 5 years were collected. At our institute,
established (13–22). For interventional CT, establish- CT-guided interventional procedures are predominantly
ing reference dose levels is a challenge because the dose (8250 of 9039; 91.3%) performed by using helical scan
for each specific procedure depends on a wide variety of modes with the operators staying in the control room dur-
factors including each patient’s size, target location, and ing radiation exposure. Therefore, this study focuses on the
the complexity of the specific interventional task (23,24). indicated CT dose, not the operators’ dose. Our hypoth-
Similarly, for fluoroscopy-guided interventional proce- esis is that by performing a large-scale survey of CT-guided
dures, because of the high individual variability of patient interventional procedures and sorting the procedures into
dose, the number of procedures recommended in the detailed clinically meaningful categories, we can achieve
CT Dose and Utilization Factors for CT-guided Interventional Procedures

To our knowledge, there is no universally agreed upon pa-

Abbreviations rameter available when reporting CTDIvol per exam by using
CTDI = CT dose index, CTDIsw = scan-length-weighted CTDIvol, the multiple CTDIvol per scan series data. The same issue also
CTDIvol = volume CTDI, DLP = dose-length product, SSDE =
size-specific dose estimate exists for SSDE per exam. In the report from American Col-
lege of Radiology dose index registry (22), the CTDIvol per
Summary exam was reported as the summation of the individual CTDI-
This study provides procedure-specific dose metric distributions and vol
from each series. The major assumption was that the exact
quantitative CT utilization factors for CT-guided interventional same anatomic region was scanned. This is not warranted for
procedures.
CT-guided interventional procedures that involve helical scans
Implication for Patient Care with different scan range. In our study, the software (Radimet-
nn Radiation dose data and CT utilization factors (average procedure- rics) reported scan-length-weighted CTDIvol and scan-length-
specific CT usage) provided in this study may be useful for quality weighted SSDE, which does not have a rigorous scientific
control of CT-guided interventional procedures and are expected definition. However, DLP as a direct measure of the total scan-
to contribute to the establishment of procedure-specific reference
radiation dose levels and CT usage levels. ner output is relatively independent to the multiple CTDIvol
challenge, and has long been adopted as the best indicator to
calculate patient effective dose for CT.
Therefore, we focused on DLP for patient dose consider-
consistent and procedure-specific radiation dose metric distribu- ation, similar to the concept of kinetic energy released per unit
tions and reference standards. Furthermore, by analyzing CT us- mass–area product (known as KAP) in fluoroscopy-guided
age-related parameters (ie, number of scan series and number of interventional procedures. Information from scan-length-
CT images), we can help create a numeric metric to evaluate and weighted CTDIvol and SSDE were included for complete re-
compare the CT usage of different subcategories of CT-guided porting. For this reason, we labeled the reported single-value
interventional procedures. CTDI and SSDE with sw in the following equations. If we
denote acquisition count (AC) as the total number of CT series
Materials and Methods occurring in one procedure and use i as the index for each CT
Our institutional review board approved this retrospective, sin- series, we can have
gle-center study (protocol number: 2015P002181) and waived AC
the requirement to obtain informed consent. Our study com- Scan Lengthtotal = ∑ Scan length(i )
plied with requirements of the Health Insurance Portability i =1 ,
and Accountability Act.
We collected dictation reports and CT radiation dose data
AC
from 9143 consecutive CT-guided interventional procedures
performed in adult patients (age at examination, 19 years;
∑CTDIvol (i ) × Scan length(i )
CTDIsw = i =1
male and female patients) from 2012 to 2017 in three dedicated Scan Lengthtotal ,
CT intervention rooms at our institute (K.Y., a medical physi-
cist with 10 years of experience, and S.G., an interventional ra-
diologist with 10 years of experience). Each room was equipped AC

with LightSpeed series CT scanners (GE Healthcare, Waukesha, ∑ SSDE (i ) × Scan length(i )
Wis) with one LightSpeed16, one LightSpeed Pro 16, and one SSDEsw = i =1

Scan Lengthtotal ,
LightSpeed Xtra. The interventional CT scans were performed at
120 kV (100 kV for intraoperative scans for chest procedures),
with two possible beam collimations (10 mm or 20 mm), three and
possible pitch values (0.563, 0.938, or 1.375), either tube cur- AC AC
rent modulation or fixed milliampere-seconds, and a recon- DLP = ∑ DLP (i ) = ∑ CTDIvol (i ) × Scan length(i )
struction section of 5 mm (2.5 mm for lung biopsy). i =1 i =1

CT scan data including CTDIvol, SSDE, scan length, DLP, = CTDIsw × Scan Lengthtotal .
acquisition count, and number of images were extracted (K.Y.)
via a commercial software platform (Radimetrics version 2.6; On the basis of the CT study description and the procedure
Bayer Healthcare, Whippany, NJ). The CTDIvol values for the dictation report, procedures were sorted (S.G. and K.Y.) into
included procedures were reported to the 32-cm body phan- the following four major interventional categories: ablation,
tom. Because CT-guided interventional procedures typically aspiration, biopsy, and drainage. The thermal ablation category
involve multiple CT scan series, such as preliminary scans, in- included cryoablation, radiofrequency ablation, and microwave
terventional scans, and postprocedure scans, the DLP and the ablation. The aspiration category included aspiration and injec-
scan length reported by the software platform (Radimetrics) tion. The biopsy category included biopsy and fiducial marker
were the summations from all the scan series. The single CT- placement. The drainage category included abdomen and chest
DIvol value reported was a scan-length-weighted CTDIvol from drainages and tube placement (eg, jejunostomy tube, gastros-
the scan series. tomy tube, and chest tube). Each major category was further

2 radiology.rsna.org n Radiology: Volume 00: Number 0— 2018

 Yang et al

Table 1: CT-guided Interventional Procedure Catego-

ries

Category Definition No. of Procedures

Ablation 679
AB1 Renal/adrenal ablation 187
AB2 Bone ablation 43
AB3 Soft tissue ablation 28
AB4 Liver ablation 394
AB5 Lung ablation 27
Aspiration 744
AS1 Abdominal aspiration 145
AS2 Chest aspiration 282
AS3 Bone aspiration/injection 257
AS4 Celiac plexus block 60
Biopsy 4425
B1 Lung biopsy 884
B2 Renal/adrenal biopsy 547
B3 Bone biopsy 760
B4 Soft tissue biopsy 253
B5 Liver biopsy 997
B6 Pancreas biopsy 25
B7 Retroperitoneal/mesenteric 959
biopsy
Drainage 2365
D1 Abdominal drainage, 1 drain 1571
D2 Abdominal drainage, 2 drains 279
D3 Abdominal drainage, 3 or 127
more drains
DC1 Chest drainage, 1 drain 354
DC2 Chest drainage, 2 drains 34
Note.—There were 8213 procedures included. All procedures
were CT guided.

divided into subcategories, either by organ (for ablation, aspira- Figure 1: Flowchart shows the process of patient
tion, and biopsy) or by number of tubes or drains (for drainage). inclusions and exclusions.
These subcategories are listed in Table 1. Our study includes a
total of 21 subcategories in the four major categories.
As shown in Figure 1, during the process of sorting, we ex- ect.org/). Distributions of six parameters including scan-length-
cluded aborted, mixed (with multiple interventional procedures), weighted CTDIvol (CTDIsw), scan-length–weighted SSDE, scan
CT fluoroscopy-guided, ultrasonography-guided, and low- length, DLP, acquisition count, and number of images were ana-
occurrence procedures (two CT myelogram procedures and four lyzed for each procedure category by the descriptive statistic out-
CT chest drainage procedures with three or more drains). Of the comes, including 25th percentile, median, and 75th percentile.
9143 procedures, 789 procedures (8.6%) were excluded due to Two sample t tests were performed to compare selected subcatego-
the use of CT fluoroscopy because the focus was on helical CT- ries to the rest of the subcategories within the same major category.
guided interventional procedures.
For the 8213 included procedures (4391 procedures per- CT Utilization Factor
formed on men and 3822 procedures performed on women), After obtaining the consistent procedure-specific distribu-
the mean patient age at examination for men was 62 years 6 15 tion of all the dose metric parameters, the last step of our
(standard deviation; age range, 19–114 years) and for women study was to derive quantitative indicators associated with the
it was 61 years 6 15 (age range, 19–113 years). There was no CT utilization of each subcategory, called procedure-specific
substantial statistical difference in age between men and women CT utilization factor. It is a single number that represents
(P = .56, paired t test). the relative use of CT guidance during a specific procedure.
A greater CT utilization factor indicates more involvement
Statistical Analysis of CT imaging guidance. With the large number of proce-
Data analysis was performed by using statistical software (R ver- dures included in each subcategory, the median value of each
sion 3.2.4; R Project for Statistical Computing, http://www.r-proj- parameter can serve as a good indicator for a representative

Radiology: Volume 00: Number 0— 2018 n radiology.rsna.org 3

CT Dose and Utilization Factors for CT-guided Interventional Procedures

scenario. By using the global median values of the related interventional procedures were observed. The lowest and
metrics from all the procedures as the baseline, a relative CT highest median DLP value ranged from 569 mGy ∙ cm
utilization factor based on a certain metric (ie, scan length (subcategory B1, CT-guided lung biopsy) to 2788 mGy
or acquisition count) for a certain procedure type was first ∙ cm (subcategory AB3, CT-guided soft tissue ablation).
derived as the ratio of median values of the sub category over The median CTDIsw value ranged from 6 mGy (subcat-
the global median as follows: egory AS2, CT-guided chest aspiration) to 20 mGy (sub-
category AS3, CT-guided bone aspiration or injection).
SL subcategory median The median scan length ranged from 399 mm (subcategory
Utilization Factor by scan length(SL ) = B3, CT-guided bone biopsy) to 2521 mm (subcategory AB3,
SL global median ,
CT-guided soft tissue ablation). The median number of images
ranged from 83 images (subcategory B4, CT-guided soft tissue
AC subcategory median biopsy) to 874 images (subcategory AB4, CT-guided lung ab-
Utilization Factor by acquisition count ( AC ) =
AC global median , lation). The median acquisition count ranged from six acquisi-
tions (subcategory B4, CT-guided soft tissue biopsy and DC1
and CT-guided chest drainage, one drain) to 27 acquisitions (subcat-
egory AB5, CT-guided lung ablation). For the most frequently
NI subcategory median
Utilization Factor by number of images ( NI ) = performed CT-guided interventional procedures within each
NI global median ,
major category, subcategories AB4 (liver ablation), AS2 (chest
aspiration), B5 (liver biopsy), and D1 (single abdominal drain),
where AC is acquisition count, NI is number of images, and the median DLP values are 2351, 657, 1175, and 1125 mGy ∙ cm,
SL is scan length. With the above definition, a utilization fac- respectively. Furthermore, for three out of these four subcat-
tor of 1 indicates the average level of CT imaging guidance egories (AS2, B5, and D1), the DLP values are significantly
across all procedure types. We chose the above three param- different from other subcategories within the same major cat-
eters to describe the CT utilization level because of their close egory (P , 2.2 3 10216, t test). Similar observations can
relationship to the involved CT guidance. We did not include be made for other parameters listed in Table 2 and Table 3.
DLP and CTDIsw to avoid the added dependency on patient These results validated the rationale to perform procedure-
size and scanning protocol. We then chose to incorporate these specific dose and utilization analysis at the subcategory level
three metrics to derive an overall utilization factor as: for CT-guided interventional procedures.
Utilization Factor by SL + Utilization Factor by AC + Utilization Factor by NI
Overall Utilization Factor =
3 . Procedure-specific CT Utilization Factors
Figure 2 plots the ranked overall utilization factor for all the
Again, a higher utilization factor indicates more CT guidance subcategories. The utilization factors ranged from 0.6 (subcate-
involvement for a specific procedure. gory B4, soft tissue biopsy) to 3.6 (subcategory AB5, lung abla-
tion). CT-guided abdominal drainages (subcategories D1, D2,
Results D3, marked with arrows on Fig 2) can provide a good example
of validation to demonstrate the clinical relevance of these de-
Procedure-specific Dose Metric and Utilization rived utilization factors. For single drain, double drains, and
Parameter Distributions three or more drains, the overall utilization factor values were
Table 2 summarizes the distribution of three CT dose related 0.9, 1.6, and 2.0, respectively.
parameters for each subcategory. The global median DLP from
our study is 1043 mGy ∙ cm. The global median CTDIsw from Discussion
our study is 12 mGy. The global median scan-length–weighted The purpose of this study is to achieve consistent and pro-
SSDE is 15 mGy and the subcategory median scan-length– cedure-specific dose metric distributions and CT utilization
weighted SSDE closely follows the corresponding CTDIsw, in- parameter distributions by performing a large-scale survey of
dicating a small deviation of patient size (typically measured as CT-guided interventional procedures and sorting the proce-
water equivalent diameter) from the 32-cm diameter phantom. dures into detailed clinically meaningful categories. Leng et al
The possible reason is the age group (62 years 6 15) and the (27) performed a similar analysis from 571 CT-guided inter-
specific patient cohort in interventional radiology. ventional procedures by using five major categories. The ad-
Table 3 summarizes the distribution of three CT utilization– dition to the literature from this study is a larger number of
related parameters for each subcategory. The global median value procedures (8213 vs 571) and more detailed categorization
of acquisition count is nine, whereas typical diagnostic abdomen (21 categories vs five categories) from a different institute.
CT scans only have two series (before and after contrast agent– The existing CTDI concept has been primarily developed
enhanced scans). The global median value of total scan length is for diagnostic CT scans, which typically contain fewer acqui-
842 mm and number of images is 166. sition series. Because of the unique use of CT guidance for
Large variations of the dose metrics and utilization pa- interventional procedures, multiple CT acquisition series (with
rameters among the different subcategories of CT-guided a median acquisition count of nine scan series from this study)

4 radiology.rsna.org n Radiology: Volume 00: Number 0— 2018

 Yang et al

Table 2: Procedure-specific Dose Metric Distributions

DLP (mGy ∙ cm) CTDIsw (mGy) Scan-length-weighted SSDE (mGy)

No. of 25th 50th 75th 25th 50th 75th 25th 50th 75th
Category Procedures Percentile Percentile Percentile Percentile Percentile Percentile Percentile Percentile Percentile
Ablation
AB1 187 1678 2364 3594 9.8 10.8 15.5 11.6 13.1 17.6
AB2 43 817 1987 3472 9.8 12.6 19.0 12.9 18.1 22.8
AB3 28 2239 2788 4293 9.7 11.2 14.0 12.1 13.7 16.6
AB4 394 1612 2351 3405 9.8 10.3 12.9 11.7 12.9 15.9
AB5 27 791 1446 2088 5.9 6.7 8.4 7.6 9.1 11.0
Aspiration
AS1 145 721 1103 1769 9.8 14.9 20.3 12.6 17.1 23.7
AS2 282 488 657 929 4.8 6.1 8.7 6.8 8.9 12.0
AS3 257 573 923 1471 14.3 19.9 25.1 19.1 25.0 32.6
AS4 60 719 1377 2107 9.8 12.3 14.5 13.3 15.7 19.2
Biopsy
B1 884 443 569 801 4.8 6.4 8.3 7.0 8.8 12.0
B2 547 762 1167 1758 9.8 13.1 19.9 11.5 15.2 22.7
B3 760 462 733 1086 14.2 18.3 23.3 19.5 25.3 31.1
B4 253 452 683 1038 12.2 18.3 23.3 17.1 23.6 32.5
B5 997 771 1175 1903 9.8 12.3 18.5 12.5 15.7 23.2
B6 25 785 1199 1713 9.8 10.9 14.3 11.9 13.9 18.3
B7 959 685 1029 1542 9.8 13.0 19.7 12.4 16.0 23.4
Drainage
D1 1571 748 1125 1866 9.8 11.7 19.0 12.0 14.8 22.8
D2 279 1514 2233 3442 9.8 15.9 22.0 12.7 18.3 25.4
D3 127 1813 2777 4281 9.8 13.2 21.6 11.4 15.6 24.8
DC1 354 697 1043 1686 9.8 9.8 18.2 11.6 14.3 22.8
DC2 34 1276 2282 3684 9.8 12.0 20.0 11.7 14.7 24.8
Overall 8213 643 1043 1798 9.8 11.6 18.8 11.5 14.9 23.2
Note.—Category definitions are provided in Table 1. CTDIsw = scan-length-weighted CT dose index, DLP = dose-length product, SSDE =
size-specific dose estimate.

are typically acquired during the same procedure. Therefore, ranking of the CT utilization for each subcategories was ob-
DLP should be considered to be a more appropriate dose served for different clinical procedures. These proposed utili-
metric for interventional CT dose analysis. Compared with zation factors could be used as reference data in many aspects
diagnostic CT, the global median CTDIsw from our study is related to interventional practice management and improve-
12 mGy, which is comparable to the achievable dose values of ment (eg, a surrogate reference to procedure complexity or
13 mGy for abdominal pelvis scans and 10 mGy for noncontrast- procedure relative value units).
enhanced chest scans as recently published by using data from In this study, we focused on CT-guided interventional pro-
American College of Radiology dose index registry (22). The cedures guided by helical scan mode. The advantages of the
global median DLP from our study is 1043 mGy ∙ cm, which use of helical scan mode include sufficient imaging coverage
is higher than the value of 639 mGy ∙ cm for abdomen and of large targeted volume and complete intervention pathway,
347 mGy ∙ cm for chest from American College of Radiology minimized occupational dose for operators, less effect from
dose index registry (22). The provided distributions of different patient motion, and consistent operation process. However,
dose metrics (CTDIsw, scan-length–weighted SSDE, DLP, scan intermittent scan mode or CT fluoroscopy mode has been cho-
length, acquisition count, and number of images) can serve as sen by many institutions with the goal to limit the radiation
a data source for the future establishment of institutional refer- field only to the small local volume related to intervention and
ence levels for CT-guided interventional procedures. to allow more so-called real-time interaction with the patient
Another contribution from our study is the development and the interventional devices (28–31). The direct comparison
of quantitative CT utilization factors for different CT-guided between helical mode and CT fluoroscopy was not the focus
interventional procedures. With the data of total scan length, of our study. However, we would like to emphasize that the
acquisition count, and number of images, we chose to use relative CT utilization factors derived from this study will be
the median value over the global median value as a surrogate useful for institutes that only use CT fluoroscopy so that they
to quantify the associated imaging guidance, and reasonable may compare between different interventional procedure types

Radiology: Volume 00: Number 0— 2018 n radiology.rsna.org 5

 6
Table 3: Procedure-specific CT Utilization Parameter Distributions

Total Scan Length (mm) Acquisition Count No. of Images

CT CT CT Overall
No. of 25th 50th 75th Utilization 25th 50th 75th Utilization 25th 50th 75th Utilization Utilization
Category Procedures Percentile Percentile Percentile Factor Percentile Percentile Percentile Factor Percentile Percentile Percentile Factor Factor
Ablation
AB1 187 1448 1902 2741 2.3 15 19 26 2.1 251 356 567 2.1 2.2
AB2 43 699 1588 2348 1.9 18 23 26 2.6 275 423 679 2.5 2.3
AB3 28 1713 2521 3428 3.0 17 23 32 2.6 321 534 866 3.2 2.9
AB4 394 1471 2050 2834 2.4 13 19 25 2.1 270 423 601 2.5 2.4
AB5 27 1500 2126 2334 2.5 23 27 38 3.0 618 874 1233 5.3 3.6
Aspiration
AS1 145 554 738 994 0.9 5 7 9 0.8 94 124 169 0.7 0.8
AS2 282 860 1083 1382 1.3 17 22 28 2.4 208 297 449 1.8 1.8
AS3 257 334 483 767 0.6 6 9 14 1.0 68 113 306 0.7 0.8
AS4 60 633 1054 1476 1.3 4 13 30 1.4 240 372 689 2.2 1.6
Biopsy
B1 884 732 938 1222 1.1 15 19 25 2.1 187 265 403 1.6 1.6
CT Dose and Utilization Factors for CT-guided Interventional Procedures

B2 547 639 818 1064 1.0 6 8 10 0.9 110 140 185 0.8 0.9
B3 760 298 399 554 0.5 6 9 12 1.0 62 89 148 0.5 0.7
B4 253 297 412 534 0.5 5 6 9 0.7 63 83 145 0.5 0.6
B5 997 642 849 1243 1.0 6 8 11 0.9 111 149 239 0.9 0.9
B6 25 753 953 1239 1.1 8 9 12 1.0 136 165 216 1.0 1.0
B7 959 582 736 957 0.9 5 7 10 0.8 99 124 168 0.7 0.8
Drainage
D1 1571 658 848 1134 1.0 5 7 10 0.8 115 152 218 0.9 0.9
D2 279 1168 1490 1977 1.8 8 12 15 1.3 202 272 377 1.6 1.6
D3 127 1537 1922 2609 2.3 9 13 19 1.4 285 367 480 2.2 2.0
DC1 354 662 840 1115 1.0 4 6 8 0.7 120 155 228 0.9 0.9
DC2 34 1244 1604 2008 1.9 8 10 15 1.1 219 293 354 1.8 1.6
Total 8213 592 842 1255 1.0 6 9 15 1.0 111 166 284 1.0 1.0
Note.—Category definitions are provided in Table 1. The CT utilization factor is defined by the ratio of the subcategory median over the global median. A higher CT utilization factor cor-
responds to a higher level of CT usage.

radiology.rsna.org n Radiology: Volume 00: Number 0— 2018

 Yang et al

Figure 2: Bar graph shows procedure-specific CT utilization factors. The procedure codes were used to label the subcategory (details in Table 3).
Each box was also color-coded per its major category. With the global reference utilization factor as 1 (average CT usage level across all proce-
dure types), CT utilization factors for abdomen drainage procedures (D1 for one drain, D2 for two drains, D3 for three or more drains) are 0.9,
1.6, and 2.0, respectively (arrows). This indicates that for procedures with three or more drains, CT usage is more than double of that for single-
drain procedures (2.0 vs 0.9).

because a more complicated procedure needs more imaging References

1. Aguirre DA, Bermudez S, Diaz OM. Spinal CT-guided interventional procedures for
guidance, independent from the type of CT utilization. For ex- management of chronic back pain. J Vasc Interv Radiol 2005;16(5):689–697.
ample, multiple drainage procedures will always require more 2. Gangi A, Dietemann JL, Mortazavi R, Pfleger D, Kauff C, Roy C. CT-guided inter-
ventional procedures for pain management in the lumbosacral spine. RadioGraphics
imaging guidance than would a single-drain procedure. 1998;18(3):621–633.
Our study had limitations. It was a single-institution study 3. Greffier J, Pereira FR, Viala P, Macri F, Beregi JP, Larbi A. Interventional spine pro-
cedures under CT guidance: How to reduce patient radiation dose without compro-
and the focus was only on helical CT scan mode, and it might mising the successful outcome of the procedure? Phys Med 2017;35:88–96.
not represent all the CT-guided interventional procedures per- 4. Haaga JR. Interventional CT: 30 years’ experience. Eur Radiol 2005;15(Suppl
4):D116–D120.
formed at other institutions. We did not analyze the data on the 5. Haaga JR, Reich NE, Havrilla TR, Alfidi RJ. Interventional CT scanning. Radiol
basis of the operator’s experience level because of the large size of Clin North Am 1977;15(3):449–456.
6. Kastler B, Fergane B. Interventional Procedures Under CT Guidance in Pain Man-
our academic institution, our long study period, and the large agement. (Cementoplasty Excluded). Interv Neuroradiol 2003;9(Suppl 2):67–73.
number of procedures that were analyzed. We believe that with 7. Kataoka ML, Raptopoulos VD, Lin PJ, Siewert B, Goldberg SN, Kruskal JB. Mul-
tiple-image in-room CT imaging guidance for interventional procedures. Radiology
the large number of procedures in each subcategory, the median 2006;239(3):863–868.
value can serve as a good representative indicator to reflect the 8. Martel Villagrán J, Bueno Horcajadas Á, Agrela Rojas E. Musculoskeletal interven-
tional radiology: ultrasound and CT. Radiologia (Madr) 2016;58(Suppl 2):45–57.
overall performance of our institute. 9. Sheafor DH, Paulson EK, Simmons CM, DeLong DM, Nelson RC. Abdominal
In conclusion, through analysis and detailed categorization percutaneous interventional procedures: comparison of CT and US guidance. Radi-
ology 1998;207(3):705–710.
of a large number of CT-guided interventional procedures, this 10. Stoffner R, Augschöll C, Widmann G, Böhler D, Bale R. Accuracy and feasibility of
study provides consistent, procedure-specific dose metric distri- frameless stereotactic and robot-assisted CT-based puncture in interventional radiol-
ogy: a comparative phantom study. Rofo 2009;181(9):851–858.
butions and quantitative CT utilization factors for CT-guided 11. Thanos L, Mylona S, Kalioras V, Pomoni M, Batakis N. Percutaneous CT-guided
interventional procedures. interventional procedures in musculoskeletal system (our experience). Eur J Radiol
2004;50(3):273–277.
Author contributions: Guarantors of integrity of entire study, K.Y., M.C.D., 12. Tsalafoutas IA, Tsapaki V, Triantopoulou C, Gorantonaki A, Papailiou J. CT-guided
interventional procedures without CT fluoroscopy assistance: patient effective dose
X.L.; study concepts/study design or data acquisition or data analysis/interpretation,
and absorbed dose considerations. AJR Am J Roentgenol 2007;188(6):1479–1484.
all authors; manuscript drafting or manuscript revision for important intellectual 13. Cho P, Seo B, Choi T, et al. The development of a diagnostic reference level on patient
content, all authors; approval of final version of submitted manuscript, all authors; dose for CT examination in Korea. Radiat Prot Dosimetry 2008;129(4):463–468.
agrees to ensure any questions related to the work are appropriately resolved, all 14. Clarke J, Cranley K, Robinson J, Smith PH, Workman A. Application of draft
authors; literature research, K.Y., X.L.; clinical studies, K.Y., S.G., M.C.D., X.L.; European Commission reference levels to a regional CT dose survey. Br J Radiol
experimental studies, K.Y., X.L.; statistical analysis, K.Y., H.Z.; and manuscript ed- 2000;73(865):43–50.
iting, K.Y., S.G., M.C.D., X.L., B.L. 15. Héliou R, Normandeau L, Beaudoin G. Towards dose reduction in CT: patient ra-
diation dose assessment for CT examinations at university health center in Canada
and comparison with national diagnostic reference levels. Radiat Prot Dosimetry
Disclosures of Conflicts of Interest: K.Y. disclosed no relevant relationships. 2012;148(2):202–210.
S.G. disclosed no relevant relationships. M.C.D. disclosed no relevant relationships. 16. Kim M, Chang K, Hwang J, Nam Y, Han D, Yoon J. Radiation Dose for Pediatric
H.Z. disclosed no relevant relationships. X.L. disclosed no relevant relationships. and Young Adult Ct: A Survey to Establish Age-Based Reference Levels of 2015-
B.L. disclosed no relevant relationships. 2016 in Korea. Radiat Prot Dosimetry 2017;175(2):228–237.

Radiology: Volume 00: Number 0— 2018 n radiology.rsna.org 7

CT Dose and Utilization Factors for CT-guided Interventional Procedures

17. MacGregor K, Li I, Dowdell T, Gray BG. Identifying Institutional Diagnostic Ref- 24. Levesque VM, Shyn PB, Tuncali K, et al. Radiation dose during CT-guided percuta-
erence Levels for CT with Radiation Dose Index Monitoring Software. Radiology neous cryoablation of renal tumors: Effect of a dose reduction protocol. Eur J Radiol
2015;276(2):507–517. 2015;84(11):2218–2221.
18. Mafalanka F, Etard C, Rehel JL, et al. Establishment of diagnostic reference levels in 25. Vano E, Järvinen H, Kosunen A, et al. Patient dose in interventional radiology: a
cardiac CT in France: a need for patient dose optimisation. Radiat Prot Dosimetry European survey. Radiat Prot Dosimetry 2008;129(1-3):39–45.
2015;164(1-2):116–119. 26. Wall B, Shrimpton P. The historical development of reference doses in diagnostic
19. Schäfer S, Alejandre-Lafont E, Schmidt T, Gizewski ER, Fiebich M, Krombach GA. radiology. Radiat Prot Dosimetry 1998;80(1-3):15–19.
Dose management for X-ray and CT: systematic comparison of exposition values 27. Leng S, Christner JA, Carlson SK, et al. Radiation dose levels for interventional CT
from two institutes to diagnostic reference levels and use of results for optimisation procedures. AJR Am J Roentgenol 2011;197(1):W97–W103.
of exposition. Rofo 2014;186(8):785–794. 28. Carlson SK, Bender CE, Classic KL, et al. Benefits and safety of CT fluoroscopy in
20. Tsapaki V, Kottou S, Papadimitriou D. Application of European Commission refer- interventional radiologic procedures. Radiology 2001;219(2):515–520.
ence dose levels in CT examinations in Crete, Greece. Br J Radiol 2001;74(885):836– 29. Nawfel RD, Judy PF, Silverman SG, Hooton S, Tuncali K, Adams DF. Patient and
840. personnel exposure during CT fluoroscopy-guided interventional procedures. Radi-
21. Verdun FR, Gutierrez D, Vader JP, et al. CT radiation dose in children: a sur- ology 2000;216(1):180–184.
vey to establish age-based diagnostic reference levels in Switzerland. Eur Radiol 30. Nishizawa K, Uruma T, Takiguchi Y, et al. Dose evaluation and effective dose estima-
2008;18(9):1980–1986. tion from CT fluoroscopy-guided lung biopsy. Igaku Butsuri 2001;21(4):233–244.
22. Kanal KM, Butler PF, Sengupta D, Bhargavan-Chatfield M, Coombs LP, Morin RL. 31. Paulson EK, Sheafor DH, Enterline DS, McAdams HP, Yoshizumi TT. CT fluoros-
U.S. Diagnostic Reference Levels and Achievable Doses for 10 Adult CT Examina- copy--guided interventional procedures: techniques and radiation dose to radiolo-
tions. Radiology 2017;284(1):120–133. gists. Radiology 2001;220(1):161–167.
23. Li J, Udayasankar UK, Carew J, Small WC. CT-guided liver biopsy: correlation of
procedure time and radiation dose with patient size, weight, and lesion volume and
depth. Clin Imaging 2010;34(4):263–268.

8 radiology.rsna.org n Radiology: Volume 00: Number 0— 2018