Key words Image Guidance - C-Arm-CT - Phantom Study - Resident in Training
Abbreviations
RiT:
Resident in training
CACT:
C-arm computed tomography
CT:
Computed tomography
DRG:
German Roentgen Society
n:
Number
p:
Level of significance
r:
Correlation coefficient
Introduction
Image-guided diagnostic and therapeutic interventions in radiology have increased
in the last 30 years [1 ]. Image guidance allows exact needle positioning, which is important for ensuring
the diagnostic significance of a biopsy and ensuring the effectiveness of local treatment
methods [2 ]
[3 ]
[4 ]. Ultrasound-guided and computed tomography-guided puncture are commonly used [5 ]. The advantage of ultrasound-guided puncture is real-time imaging. The disadvantages
include a low penetration depth, particularly in the case of obesity or the superimposition
of air, and the dependence on the operator [5 ]. The CT-guided puncture technique benefits from operator-independent, three-dimensional
image information. However, real-time information about the progression of the puncture
needle is not available without navigation or is only available on a limited basis
in the case of CT fluoroscopy [5 ]
[6 ]. There are various options for performing CT-guided interventions. On the one hand,
operators can leave the CT room or can use radiation protective equipment and remain
next to the CT gantry, while computed tomography with minimal slices (typically 3
slices with a slice thickness of 5 mm) focused on the puncture tract is performed
repeatedly (“quick-and-check”). When the operator leaves the CT room, there is no
radiation exposure. If the examiner remains next to the gantry, the radiation exposure
will be negligible. However, the puncture needle must be advanced sequentially and
without real-time imaging. On the other hand, real-time imaging is possible with CT
fluoroscopy. Using radiation protective equipment, the operator remains in the room.
Depending on the technique, the operator’s hand even remains on the needle, which
is associated with radiation exposure [7 ].
An alternative to conventional CT is C-arm computed tomography (CACT). The advantage
of this puncture method is the combination of spatial 3 D CACT information with real-time
fluoroscopy information, possibly with the overlay of trajectories [8 ]. This method reduces the radiation dose [9 ] and could also make image guidance of complex, double-angulated puncture paths easier
compared to CT [10 ]. The literature specifies a reduction of the effective patient dose for CACT of
up to 40 % compared to conventional CT-guided puncture [9 ]. Depending on the study [10 ]
[11 ]
[12 ]
[13 ]
[14 ]
[15 ]
[16 ], the CACT-guided puncture method even seems to be superior to conventional methods
like CT-guided puncture with regard to puncture accuracy.
The puncture deviation and puncture duration of CACT-guided puncture methods performed
by experienced interventional radiologists were recently examined by Busser et al.
in a phantom study [10 ]. The training and experience of residents in training (RiT) in radiology with CT-guided
and CACT-guided puncture have not yet been studied. However, studies show that the
simulation of image-guided methods can improve the learning curve for vascular interventions
among RiTs [17 ]
[18 ]. The goal of our study was therefore to compare CT image guidance and CACT image
guidance among RiTs with limited interventional experience based on puncture deviation
and duration in a phantom with different degrees of spatial complexity and to correlate
the puncture deviation and duration with the RiTs’ manual and spatial skills.
Materials and Methods
Study participants and covariates
As part of the structured program “Researchers for the Future” created in 2010 for
the targeted promotion of young radiologists by the German Roentgen Society, the Conference
of Professors of Radiology, and the Academy for Further and Continuing Education in
Radiology, 38 RiTs from university hospitals in Germany and Austria were invited to
the Hannover Medical School on March 14 and 15, 2019. 35 RiTs attended. Five RiTs
did not actively participate due to organizational reasons and another five due to
personal reasons. Thus, a total of 25 RiTs performed punctures in the phantom.
Prior to the event, information regarding the RiTs’ level of training was recorded
using a questionnaire. The questionnaire included questions regarding professional
experience in radiology in years and the number of independently performed puncture
procedures (ultrasound-guided, CT-guided), angiography procedures, and CACT-guided
puncture procedures. Moreover, the questionnaire included the self-assessment of their
manual and spatial skills on a scale of 1–6 (1: very good, 2: good, 3: satisfactory,
4: sufficient, 5: deficient, 6: unsatisfactory) and a qualitative and quantitative
assessment of their experience playing a musical instrument, video games, and ball
games (type of musical instrument, video game console, and ball sport as well as the
number of years of experience).
After a short training session in CT-guided and CACT-guided puncture techniques on
the phantom, the RiTs were divided into 6 equal groups with comparable radiology experience
to perform the puncture procedures. Two puncture procedures either with CT or CACT
image guidance were planned for every RiT a time interval of 30 minutes.
Puncture phantom
Puncture phantoms were used to analyze the puncture deviation. A three-dimensional
printed model with one entry ring and six target rings made of resin (Form 2, clear
resin, Formlabs, Somerville, Massachusetts, USA) is embedded in a gelatin matrix (4
liters of distilled water, 350 grams of 7 % gelatin, 35 grams of flour, and 15 milliliters
20 % chlorhexidine) ([Fig. 1 ], [2 ]). After the end of the puncture procedure, the target position was marked with a
5-millimeter guidewire fragment (Transend Shapeable Tip, Guidewire with ICE Hydrophilic
Coating, 190 cm, 0.014 inch, < 0.37 mm; Boston Scientific, Marlborough, Massachusetts,
USA) that was advanced through the puncture needle (one-piece angiographic needle
with snap-on wing, 18 gauge, 70 mm, 0.038 inch; Cordis, Santa Clara, California, USA/
Chiba Access and Biopsy Needle, 22 gauge, 15 cm; COOK MEDICAL, Bloomington, Indiana,
USA). A total of 12 puncture phantoms were available for the 6 groups. After completion
of all CT-guided and CACT-guided puncture procedures, the positions of the wire markers
in the phantom were detected with a native CT scan (helical, 271 slices, 1.25 mm slice
thickness, 120 kV, 10 mA; GE Lightspeed 16; General Electric, Boston, Massachusetts,
USA). The shortest distance from the distal end of the wire marker to the center of
the target ring on CT (puncture deviation [mm]) was measured with a ruler function
(Visage 7, Visage Imaging GmbH, Berlin, Germany). In addition, the needle placement
time (puncture duration [min]) from the start of the first CT or CACT scan to the
successful positioning of the wire marker was documented.
Fig. 1 Phantom. This figure shows a photograph of the phantom. The phantom has an entry
ring and six target rings made of clear resin and was placed in a non-radiopaque gelatin
matrix.
Fig. 2 Computed tomography-guided and C-arm computed tomography-guided puncture. a The first puncture was planned in a transaxial or single-angulated needle path, shown
as an example with CT image control. b The second puncture was carried out in a complex, single-angulated, or double-angulated
needle path as shown with CT guidance. c The fluoroscopic image shows the top view of the puncture needle, which is located
within the red labeled crosshair of the navigation software (“bulls eye view”). d In the lateral view, the entire puncture needle is shown in the fluoroscopic image
and the needle path of the navigation software is labeled in green (“progression view”).
Needle placement method
The puncture was performed either with CT guidance (GE Lightspeed 16; General Electric
Healthcare, Chicago, Illinois, USA) or CACT guidance (Siemens Pheno; Siemens Healthineers,
Erlangen, Germany). The first puncture was performed with a transaxial needle path
and the second puncture with a single angulated needle path or the first puncture
was performed with single angulated and the second puncture with a double angulated
needle path (see [Fig. 2 ]).
CT-guided puncture
At the start, a native CT scan (helical, 271 slices, 1.25 mm slice thickness, 120 kV,
10 mA) of the puncture phantom with conventional, radiopaque markers was acquired.
The optimal entry point and the needle path to the target were determined. The marker
was removed and after placement of the needle at the point of entry, native CT scans
in a transaxial direction were repeatedly acquired to check the position of the tip
of the needle (transaxial, 5 slices, 2.5-mm slice thickness, 120 kV, 60 mA).
CACT-guided puncture
For the CACT-guided puncture, the acquisition of a native CACT scan (5 s, 95 projections/s,
397 projections, 90 kV, 100 mA) and reconstruction of a three-dimensional dataset
were conducted. The entry point as well as the target point were determined by the
person performing the puncture using navigation software. The needle path was calculated
automatically. In the first step, the C-arm was automatically positioned in a projection
plane perpendicular to the direction of puncture (“bulls eye view”) (see [Fig. 2 ]). The intersecting planes of the laser cross hairs integrated in the detector of
the angiography system mark the entry point on the phantom and the trajectory. To
monitor the progression of the puncture needle in real time via fluoroscopy, the C-arm
was automatically moved to a projection plane parallel to the planned needle path
(“progression view”) (see [Fig. 2 ]). Both view settings could be changed by each resident as needed until the needle
or the marker was placed in the target ([Fig. 3 ]).
Fig. 3 Computed tomography of the phantom. This three-dimensional reconstruction of the
native CT scan shows the radiopaque markers, which were positioned via CT-guided and
C-arm CT-guided puncture. Two paper clips were embedded in the gelatin matrix as additional
radiopaque markers for spatial orientation.
Evaluation of the phantom study
In a subsequent questionnaire using SurveyMonkey (www.surveymonkey.com , SurveyMonkey Inc., San Mateo, California, USA) the 35 RiTs who were present at the
Hannover Medical School were invited to evaluate the phantom study (10 of the RiTs
did not actively perform any puncture procedures). The following questions were answered
using a Likert scale from 1–5 (1: completely disagree, 2: disagree, 3: neither agree
nor disagree, 4: agree, 5: completely agree):
Is the phantom generally suitable for CT/CACT-guided puncture training?
Can training on a phantom improve patient care?
Should CT/CACT-guided puncture training on a phantom be part of the RiT program?
Is the currently offered training (e. g., at conventions, in workshops, or in your
own department) regarding CT/CACT-guided interventions in Germany and Austria sufficient
(prior to the pandemic)?
Statistical analysis
The information provided by the RiTs in the questionnaire was recorded in the categories
described above with the mean value and standard deviation. The level of training
of the RiTs who performed puncture with CT image guidance was compared with that of
the RiTs who performed puncture with CACT image guidance. The puncture deviation and
puncture duration were compared between the methods and between the first and second
puncture. The puncture deviation and puncture duration were then correlated with the
self-assessment regarding manual and spatial skills to detect a potential difference
and any advantage for learning image-guided methods. The evaluation results were documented
with the number of responses on the Likert scale.
The statistical evaluation was performed with the R 3.6.2 statistical computation
system (https://www.r-project.org ). In the case of non-parametric distribution analyzed with the Shapiro-Wilk test,
the Mann-Whitney U test for independent samples was used for the comparison between
CT and CACT image guidance. The Wilcoxon test for independent samples was performed
for the comparison between the first and second puncture within a group. One participant,
who only performed the first puncture in the available time, was excluded from the
independent comparison between the first and second puncture within the group with
CT image guidance. The correlation was analyzed with the Spearman rank correlation
coefficient (r). Two-sided testing was performed with a significance level of p < 0.05.
Results
Study participants and covariates
The average professional experience in radiology per RiT was 3 ± 1 year. The number
of already performed puncture procedures per RiT was 14 ± 34 for ultrasound-guided
puncture, 36 ± 44 for CT-guided puncture, 30 ± 70 for angiography, and 8 ± 31 for
CACT-guided puncture. In the self-assessment, both manual and spatial skills were
assigned a value of 2 ± 1. 18 RiTs had experience playing a musical instrument, 20
playing video games, and 17 playing ball sports. The accordion, cello, electric bass,
guitar, clarinet, piano, organ, German flute, and violin were listed as the musical
instruments. Basketball, soccer, handball, squash, tennis, table tennis, and volleyball
were specified as the types of ball sport. Neither the level of training nor the experience
with punctures was statistically different between the RiTs that performed CT-guided
puncture and those that performed CACT-guided puncture ([Table 1 ]).
Table 1
Residents in training in radiology.
CT
(n = 11)
CACT
(n = 14)
p-value
professional experience in radiology (years)
3 ± 1
3 ± 1
1
number of conventional CT-guided puncture procedures
36 ± 47
35 ± 42
0.934
number of ultrasound-guided puncture procedures
14 ± 25
14 ± 40
0.483
number of angiography procedures
11 ± 19
45 ± 89
0.415
number of CACT-guided puncture procedures
5 ± 15
11 ± 39
0.466
This table shows the professional experience of residents in training in radiology
who performed either computed tomography (CT)-guided or C-arm computed tomography
(CACT)-guided puncture. The mean values and standard deviation as well as the p-value
of the Mann-Whitney-U test are given. Abbreviations: n = number of residents in training.
Puncture deviation in the phantom
The difference in the puncture deviation between CT and CACT was not significant (7.2
± 3.3 mm and 7.9 ± 3.3 mm) (p = 0.337). There was also no statistical difference between
the first and second puncture in the CT group (6.4 ± 2.7 mm and 8.5 ± 3.5 mm; p = 0.130)
and in the CACT group (8.3 ± 4.2 mm and 7.6 ± 2.2 mm; p = 0.391). The results are
provided in detail in [Table 2 ], [3 ].
Table 2
Target deviation and puncture duration between computed tomography-guided and C-arm
computed tomography-guided puncture.
CT
(n = 21)
CACT
(n = 28)
p-value
puncture deviation [mm]
7.2 ± 3.3
7.9 ± 3.3
0.337
puncture duration [min]
11 ± 11
6 ± 2
< 0.001
This table lists the target deviation and puncture duration by the residents who performed
either computed tomography (CT)-guided or C-arm computed tomography (CACT)-guided
puncture. Mean values and standard deviation as well as the p-value of the Mann-Whitney-U
test are given. Abbreviations: min = minute(s), mm = millimeter(s) and n = number
of values.
Table 3
Target deviation and puncture duration between the first and second puncture of the
computed tomography-guided or C-arm computed tomography-guided puncture.
first puncture
second puncture
p-value
CT
n = 10
n = 10
puncture deviation [mm]
6.4 ± 2.7
8.5 ± 3.5
0.130
puncture duration [min]
13 ± 17
9 ± 3
0.719
CACT
n = 14
n = 14
puncture deviation [mm]
8.3 ± 4.2
7.6 ± 2.2
0.391
puncture duration [min]
7 ± 2
5 ± 2
0.006
This table shows the target deviation and puncture duration for the first and second
puncture by the residents who performed either computed tomography (CT)-guided or
C-arm computed tomography (CACT)-guided puncture. Mean values and standard deviation
as well as the p-value of the Wilcoxon test are shown. Abbreviations: min = minute(s), mm = millimeter(s)
and n = number of values.
Puncture duration on the phantom
The puncture duration of CACT-guided puncture (6 ± 2 min) was significantly shorter
than that of CT-guided puncture (11 ± 11 min) (p < 0.001). In the case of CACT, the
second, more difficult puncture was performed more quickly (5 ± 2 min) than the first
puncture (7 ± 2 min) (p = 0.006). In the case of CT-guided puncture, there was no
statistical difference between the first and second puncture (13 ± 17 min compared
to 9 ± 3 min) (p = 0.719). The results are shown in [Table 2 ], [3 ].
Influence of the self-assessment
The self-assessment of manual skills did not correlate with the puncture deviation
(r: + 0.271; p = 0.059) and the puncture duration (r: –0.204; p = 0.158). There was
a significant correlation between the self-assessment of spatial skills and puncture
deviation (r: –0.089; p = 0.541) but not between spatial skills and puncture duration
(r: –0.089; p = 0.541). The results are shown in [Table 4 ].
Table 4
Impact of self-assessment on target deviation and puncture duration in the phantom.
correlation
r-value
(n = 49)
p-value
puncture deviation/spatial skills
+ 0.356
0.011
puncture deviation/manual skills
+ 0.271
0.059
puncture duration/spatial skills
–0.089
0.541
puncture duration/manual skills
–0.204
0.158
The table shows the correlation between target deviation and puncture duration with
self-assessment of spatial skills and manual skills of the residents in training.
The correlation coefficient, r-value, and p-value of the Spearman rank correlation
analysis are shown. Abbreviations: n = number of values.
Evaluation of the phantom study
Willingness to participate in the subsequent questionnaire regarding the phantom study
was high (33 of 35 RiTs (94 %)). In total, 97 % of the RiTs agreed that the phantom
is generally suitable for CT/CACT-guided puncture training (number of responses on
the Likert scale: 1 = 0, 2 = 0, 3 = 1, 4 = 14, 5 = 18) and that patient care can be
improved by training on a phantom (Likert scale response distribution: 1 = 0, 2 = 0,
3 = 1, 4 = 10, 5 = 22). 91 % of the RiTs found that CT/CACT-guided puncture training
on a phantom should be part the RiT program (Likert scale response distribution: 1 = 0,
2 = 2, 3 = 1, 4 = 5, 5 = 25). In contrast, 27 % of the RiTs neither agreed nor disagreed
and 61 % disagreed that the currently offered CT/CACT-guided intervention training
in Germany and Austria is sufficient (Likert scale response distribution: 1 = 5, 2 = 15,
3 = 9, 4 = 2, 5 = 2).
Discussion
In our study, RiTs from university radiology departments from all over Germany participating
in the “Researchers for the Future” program of the German Roentgen Society performed
puncture procedures. On average, the RiTs selected by the individual university hospitals
were in the third year of their RiT program and had already performed the number of
non-vascular interventions required by the Specialty Training Regulations [19 ]. Based on clinical practice, experience with CT-guided puncture is greater than
with CACT-guided puncture as expected. Overall, a slightly greater puncture deviation
from the target of approx. 7 mm was seen in our study in the CT group and the CACT
group compared to the literature, e. g. a deviation of 3 mm among experienced interventional
radiologists was reported in the phantom study by Busser et al. and between 3 mm and
12 mm in clinical practice [10 ]
[20 ]
[21 ]. However, since the RiTs performed puncture procedures after a brief introduction
to an unfamiliar environment in our phantom study, the puncture deviation is not unexpected
and is clinically acceptable in many cases.
The puncture time for CACT-guided puncture (6 ± 2 min) was significantly shorter than
for CT-guided puncture (11 ± 11 min.) This could be due to the workflow since the
RiTs wore radiation protective clothing and remained in the angiography room during
CACT-guided puncture while the RiT of the CT group left the examination room during
CT-guided puncture. Moreover, CACT puncture guidance is supported by a navigation
tool while no software support was available for CT. This navigation software seems
to be intuitive even for people with minimal experience performing puncture procedures
since the second and significantly more difficult CACT-guided puncture was faster
than the first puncture. This learning effect was lower and not statistically significant
in the CT group without a navigation tool. The learning effect regarding puncture
deviation was also seen in the phantom study by Busser et al. even among experienced
interventional radiologists [10 ]. Therefore, our study shows that navigated CACT-guided puncture allows a steep learning
curve even among inexperienced RiTs in radiology. CT-guided puncture would presumably
also benefit from software support provided that its use is similarly intuitive. Hence,
CT-guided and CACT-guided interventions using modern navigation tools should be simulated
and tested on phantoms and be offered on a more intensive basis than a training unit
during the RiT program to prepare RiTs with minimal experience for clinical application.
There was no significant correlation between the self-assessment of manual skills
and the puncture deviation or puncture duration. There was also no correlation between
the self-assessment of spatial skills and the puncture duration. However, there was
a moderate positive correlation between spatial skills and puncture deviation. The
last result highlights the relevance of spatial skills for learning image-guided interventions.
Although this skill can be different in people, it can be improved by training [22 ]. Therefore, spatial skills training, for example, using a phantom or simulator could
contribute to a steeper learning curve for image-guided interventions. This has already
been shown by other studies for endovascular interventions. For example, the fluoroscopy
time and the intervention duration of subsequent interventions in the clinical routine
could be significantly reduced by simulator training, e. g. stent implantation in
the internal carotid artery or diagnostic coronary angiography [17 ]
[18 ]. Since most RiTs had two or more hobbies requiring manual skills, a further statistical
evaluation was not possible due to the lack of a group of RiTs without hobbies.
The questionnaire regarding the phantom study had a high response rate (94 %) and
a uniform response pattern resulting in a clear result. The phantom used in the study
was rated as suitable for CT/CACT-guided puncture training and the RiTs felt that
the training on a phantom could also help to improve patient care. Even though training
on a phantom is not currently part of the RiT program, the RiTs felt that it should
be as in other professional groups like pilot training. The RiTs consider the training
options for non-vascular interventions on a phantom currently available in Germany
and Austria as insufficient.
Our phantom study has some limitations. The assessment of manual and spatial skills
was subjective, but could be provided objectively by tests. The use of the “quick-and-check”
technique for CT and the navigation software for CACT guidance limits the ability
to compare the methods even though this corresponds to the clinical routine since
tools are not used at many institutions for CT-guided biopsies and drainage procedures.
In contrast, the CACT-guided technique is hardly feasible without navigation. Moreover,
the radiation exposure in CACT-guided puncture techniques with overlay is 40 % lower
compared to conventional CT-guided puncture [9 ]. The last and most important limitation is the low number of experiments and participants.
Although the average puncture experience of the RiTs corresponded to their level of
training, the group was very heterogenous and that explains the high deviation of
values and corresponding limitations of the statistical analysis. Unfortunately, the
time in the “Researchers for the Future” program was limited so that the number of
puncture procedures could not be increased and the RiTs did not have time to perform
the two puncture techniques. Our results and the questionnaire can be used as the
basis for further studies with a corresponding number of cases and study design in
order to improve simulators and at the same time to further evaluate the advantages
and disadvantages of puncture methods, particularly for those with limited experience
performing puncture procedures.
Funding
Deutsche Röntgengesellschaft e. V. (Forscher-für-die-Zukunft)
Clinical relevance of the study
In the group of RiTs selected from university radiology departments as part of the
“Researchers for the Future” program, the experience with CT-guided puncture corresponds
to the standard defined by the Specialty Training Regulations.
Although experience with CACT image guidance is significantly lower, CACT with software
support seems to have a steeper learning curve than conventional CT.
The RiTs rated their skills high and achieved accuracy in the study corresponding
to their level of training.
