ORIGINAL ARTICLE
The Short Breath-Hold Technique, Controlled Aliasing in Parallel
Imaging Results in Higher Acceleration, Can Be the First Step to
Overcoming a Degraded Hepatic Arterial Phase in Liver Magnetic
Resonance Imaging
A Prospective Randomized Control Study
Jung Lim Yoo, MD, Chang Hee Lee, MD, PhD, Yang Shin Park, MD, Jeong Woo Kim, MD, Jongmee Lee, MD,
Kyeong Ah Kim, MD, Hae Young Seol, MD, and Cheol Min Park, MD
Objective: The aim of this study was to assess whether a short breath-hold technique can improve hepatic arterial phase (HAP) image quality in gadoxetic acid–
enhanced magnetic resonance (MR) imaging compared with a conventional long
breath-hold technique.
Materials and Methods: Institutional review board approval and patient consent
were obtained for this prospective randomized control study. One hundred nineteen patients undergoing gadoxetic acid–enhanced MR imaging were randomly
assigned to groups A or B. Group A patients underwent an 18-second long
breath-hold MR technique (conventional VIBE [volumetric interpolated breathhold examination] technique with GRAPPA [generalized autocalibrating partially
parallel acquisition]), and group B patients underwent a 13-second short breathhold MR technique (VIBE technique with CAIPIRINHA [controlled aliasing in
parallel imaging results in higher acceleration]). Respiratory-related graphs of
the precontrast and HAP were acquired. The breath-hold degree was graded
based on the standard deviation (SD) value of respiratory waveforms. Gadoxetic
acid–related dyspnea was defined as when the SD value of the HAP was 200
greater than that of the precontrast phase without degraded image quality in the
portal and transitional phases (SD value of the HAP − SD value of the precontrast
phase). The overall image quality and motion artifacts of the precontrast and HAP
images were evaluated. The groups were compared using the Student t or Fisher
exact test, as appropriate.
Results: The incidence of breath-holding difficulty (breath-hold grades 3 and
4) during the HAP was 43.6% (27/62) and 36.8% (21/57) for group A and B, respectively. The SD value during the precontrast phase and the SD value difference
between the precontrast and HAP were both significantly higher in group A than
in group B (P = 0.047 and P = 0.023, respectively). Gadoxetic acid–related dyspnea was seen in 19.4% (12/62) of group A and 7.0% (4/57) of group B. Group B
showed better precontrast and HAP image quality than group A (P < 0.001). Degraded HAP (overall image quality ≥4) was observed in 9.7% (6/62) and 3.5%
(2/57) of group A and B, respectively.
Conclusions: The short breath-hold MR technique, CAIPIRINHA, showed better HAP image quality with less degraded HAP and a lower incidence of breathhold difficulty and gadoxetic acid–related dyspnea than the conventional long
breath-hold technique.
Received for publication November 20, 2015; and accepted for publication, after
revision, November 29, 2015.
From the Department of Radiology, Korea University Guro Hospital, Korea University College of Medicine, Seoul, South Korea.
Conflicts of interest and sources of funding: Supported by a research grant from Bayer
Korea Ltd (ISS#17153).
The authors report no conflicts of interest.
Supplemental digital contents are available for this article. Direct URL citations appear
in the printed text and are provided in the HTML and PDF versions of this article
on the journal's Web site (www.investigativeradiology.com).
Correspondence to: Chang Hee Lee, MD, PhD, Department of Radiology, Korea University Guro Hospital, Korea University College of Medicine, 80 Guro-dong,
Guro-gu, Seoul 152-703, South Korea. E-mail: [email protected].
Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
ISSN: 0020-9996/16/0000–0000
DOI: 10.1097/RLI.0000000000000249
Investigative Radiology • Volume 00, Number 00, Month 2016
Key Words: gadoxetic acid, liver, magnetic resonance imaging, breath-hold,
hepatic arterial phase, dyspnea
(Invest Radiol 2016;00: 00–00)
T
he hepatic arterial phase (HAP) in liver magnetic resonance imaging (MRI) is a critical phase essential for detection and characterization of focal liver lesions.1–4 The image quality of the HAP must be
optimal without artifacts to accurately diagnose focal liver lesions. Recently, gadoxetic acid–enhanced MRI has been widely used for liver
imaging as it offers standard dynamic images as well as hepatobiliary
phase images within only 20 minutes.2,3,5–8 However, degraded HAP
imaging has been more frequently reported with contrast-enhanced MRI
using gadoxetic acid than with other gadolinium contrast agents.1,6,9–11
There have been previous reports on gadoxetic acid–related acute transient dyspnea that may cause respiratory motion-related artifacts in the
HAP, leading to suboptimal or nondiagnostic HAP imaging.6,9–11 According to previous reports, the incidence of “transient severe motion” in the
HAP ranges from 4.8% to 18.3%.6,9–13 Dyspnea, whether induced by
gadoxetic acid or as the result of breathlessness due to long breathhold time, may disturb breath-holding and degrade HAP image quality.
Therefore, reducing breath-hold time during the HAP may be crucial in
reducing dyspnea and motion artifacts resulting in improved HAP
image quality.
Evaluating breath-hold degree may be important for maintaining optimal HAP image quality because poor breath-holding can
cause degradation of image quality.14 In a previous study, breathholding difficulty was evaluated directly and objectively by assessing
respiratory-related graphs, and the breath-hold degree correlated with
the overall image quality and motion artifacts.14 Therefore, evaluating
the degree of breath-hold objectively and directly by analyzing respiratory patterns may be useful for image quality analysis. To the best of our
knowledge, no other study has prospectively evaluated differences in
image quality and respiratory patterns between short and long breathhold MR techniques.
The purpose of this prospective randomized control study was to
assess whether a short breath-hold technique can improve HAP image
quality in gadoxetic acid–enhanced MRI compared with a conventional
long breath-hold technique and also to objectively evaluate if shortening
breath-hold time can reduce gadoxetic acid–related respiratory difficulty by evaluating respiratory-related graphs.
MATERIALS AND METHODS
This prospective observational study was approved by the institutional review board. Before the investigation, the adequate sample size was estimated as 128 patients using the independent z-test.
We verbally explained to the patients the purpose and method of this
study and emphasized that there was no additional harm in participating in the study as they would receive the standard amount of
www.investigativeradiology.com
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
1
Investigative Radiology • Volume 00, Number 00, Month 2016
Yoo et al
FIGURE 1. Flowchart of the study population. Among the 128 patients enrolled in this study, 9 patients (2 from group A and 7 from group B) were
excluded because the respiratory-related graphs could not be evaluated for the following reasons: (1) absence of SD value (n = 1), (2) linear waveform
(n = 3), (3) bizarre waveform (n = 3), and (4) minimum SD value of zero (n = 2). Reasons 1 and 2 were caused by technical failure (disconnected electrical
cord or bluetooth device); reason 3 was induced by coughing or sneezing; and reason 4 was due to physical movement of the patient. As a result, a total
of 119 patients (62 in group A and 57 in group B) were included in this study. Figure 1 can be viewed online in color at www.investigativeradiology.com.
contrast media. After thorough review, the patients signed the informed consent forms.
of contrast media. The patients were instructed to hold their breath at
the end of inspiration, and then arterial phase acquisition was initiated.
Subjects
Image Analysis
One hundred twenty-eight eligible patients (89 men and 39
women; mean [SD] age, 56.1 [12.2] years) were prospectively enrolled
from October 1, 2013 to June 30, 2015 at a single tertiary university
hospital. Patients were scheduled to undergo liver MRI for hepatocellular carcinoma (n = 63), metastasis (n = 2), cholangiocarcinoma (n = 2),
liver cirrhosis (n = 21), hemangioma (n = 8), focal nodular hyperplasia
(n = 6), focal eosinophilic infiltration (n = 3), fatty liver (n = 12), arterioportal shunt (n = 5), and hepatic cyst (n = 6).
The 128 patients were randomly assigned to groups A and B.
Group A patients underwent an 18-second long breath-hold MR technique, and group B patients underwent a 13-second short breath-hold
MR technique. Among them, 9 patients (2 from group A and 7 from
group B) were excluded because the respiratory-related graphs
could not be evaluated (Fig. 1). As a result, a total of 119 patients (62 in
group A and 57 in group B) were included in this study (Supplement
E1, Supplemental Digital Content 1, http://links.lww.com/RLI/A248).
For qualitative analysis, 2 abdominal radiologists (J.L.Y and
Y.S.P., with 3 and 7 years of abdominal imaging experience, respectively) blinded to the MR protocol and breath-hold pattern independently reviewed the precontrast and HAP images of both groups.
The overall image quality and motion artifacts were assessed. The
overall image quality was scored as follows with the highest grade
indicating worst image quality: (1) excellent, (2) good, (3) acceptable, (4) poor, and (5) unreadable.14 Motion artifacts were graded
Magnetic Resonance Imaging
Liver MRI was performed on a 3 T MR scanner (MAGNETOM
Skyra; Siemens, Erlangen, Germany) with a standard 18-channel body
matrix coil and a table-mounted 32-channel spine matrix coil. Specific
MR sequence parameters for both groups are summarized in Table 1.
For dynamic imaging, a fat-suppressed T1-weighted 3D GRE
volumetric interpolated breath-hold examination (VIBE) sequence
was obtained using generalized autocalibrating partially parallel acquisition (GRAPPA) in group A and controlled aliasing in parallel imaging
results in higher acceleration (CAIPIRINHA) in group B. After the
precontrast scan, intravenous bolus of 0.1 mL/kg gadoxetic acid
(Primovist; Bayer Pharma AG, Berlin, Germany) was administered
at a rate of 1 mL/s, followed by 25-mL saline flushing. Under real-time
fluoroscopic monitoring, the arterial phase was obtained immediately after detection of the contrast media in the descending aorta
for group A and in the aortic arch for group B.1 The portal venous,
transitional, and hepatobiliary phase images were performed at
70 seconds, 2 minutes, and 20 minutes, respectively, after injection
2
www.investigativeradiology.com
TABLE 1. MR Protocol and Sequence Parameters of Group A and B
MR protocol
TR, ms
TE, ms
Flip angle,
degrees
Slice thickness, mm
Matrix
Field of view
Voxel size, mm3
Acceleration factor
Delta shift
Acquisition time, s
K-space
Group A (n = 64)
Group B (n = 64)
VIBE sequence
with GRAPPA
4.12
1.20
9
VIBE sequence
with CAIPIRINHA
4.12
1.89
9
3
416 324
316 319
0.76 0.98 3
2
NA
18
Linear
3
384 191
360 363
0.94 1.90 3
4 (2 2)
1
13 (3 s calibration
time + 10 s
acceleration time)
Segmented linear
MR indicates magnetic resonance; TR, repetition time; TE, echo time; VIBE,
volumetric interpolated breath-hold examination; GRAPPA, generalized autocalibrating partially parallel acquisition; CAIPIRINHA, controlled aliasing in
parallel imaging results in higher accelerations; NA, not applicable.
© 2016 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
Investigative Radiology • Volume 00, Number 00, Month 2016
as follows: (1) no artifacts, (2) mild artifacts without diagnostic impairment, (3) moderate artifacts without diagnostic impairment, (4) severe artifacts with diagnostic impairment, and (5) nondiagnostic.6
The scores for each image quality parameter between the 2 readers
were averaged to obtain the mean score, which was used in further
analysis. Mean overall image quality scores of 4 or higher indicated
degraded HAP.
Respiratory-Related Graphs
Before the MR examination, a pneumatic cushion usually used
for patient monitoring was placed on the epigastric area of each patient
and fastened with a body coil. The cyclic expansion or contraction of
the thorax was transmitted to the pneumatic cushion, and a respiratory
curve based on the pressure change was acquired. The respiratory signals were recorded during the precontrast and HAP and plotted on a
graph. The y axis of the graph represented pressure change caused by
respiratory movement, measured in arbitrary units, while the x axis
represented time.14
Respiratory Pattern Analysis
For quantitative analysis, an experienced abdominal radiologist
(C.H.L., with 17 years of abdominal imaging experience) blinded to
the MR protocol and image quality scores analyzed the respiratory patterns during the breath-holding period by assessing the respiratoryrelated graph acquired during the precontrast and HAP.
According to the method of a previous study,14 the breath-hold
degree was graded based on SD value with the highest grade indicating
the worst breath-hold: grade 1, excellent breath-hold (SD <100); grade
2, mild difficulty in breath-holding (SD 100–200); grade 3, moderate
difficulty in breath-holding (SD 200–300); and grade 4, severe difficulty in breath-holding (SD >300; Fig 2). Patients with breath-hold
grades 3 and 4 (SD >200) were considered to have respiratory difficulty.
When the SD value during the HAP was 200 greater than that of
the precontrast phase without degraded image quality in the portal and
transitional phases, it was defined as gadoxetic acid–related dyspnea
(GARD; SD value of the HAP − SD value of the precontrast
phase; Fig. 3).14
Survey Questionnaire
Immediately after MR examination, patients blinded to the MR
protocol underwent a survey questionnaire to investigate the subjective
feeling of dyspnea. They were asked if they felt they had held their
breath well after contrast media administration; if they did not hold their
breath well, they were asked if they ordinarily have difficulty holding
their breath.
Statistical Analyses
The SD values of the precontrast and HAP and the SD value differences between both phases were compared between groups A and B
using the Student t test. Differences in the incidence of each breath-hold
grade, breath-holding difficulty, GARD, and degraded HAP between
the 2 groups were assessed using the Fisher exact test.
Interobserver agreement of the image quality scores for the
precontrast and HAP was assessed by calculating intraclass correlation
coefficients. The mean image quality scores of patients in groups A and
B, and patients with and without GARD were compared using the Student t test.
P value less than 0.05 was considered statistically significant.
Statistical analyses were performed with statistical software (SPSS
version 20.0; IBM, Armonk, NY).
RESULTS
The patient demographics and characteristics are presented in
Supplement E2, Supplemental Digital Content 2, http://links.lww.com/RLI/A249.
© 2016 Wolters Kluwer Health, Inc. All rights reserved.
Overcoming Degraded MR Hepatic Arterial Phase
Image Analysis
Reader agreement was good to excellent for all image quality
scores in both the precontrast and HAP, with intraclass correlation coefficients of 0.707 to 0.881. The mean image quality scores of the
precontrast and HAP in both groups are summarized in Table 2 and Supplement E3, Supplemental Digital Content 3, http://links.lww.com/RLI/A250.
The mean overall image quality and motion artifact scores of group A
were significantly higher than group B in both phases (all P < 0.001).
In other words, group B demonstrated better image quality. None of
the patients in either group showed an overall image quality score of
5 (worst image quality) in either the precontrast or HAP. Degraded
HAP (overall image quality ≥ 4) were seen in 8 patients in the total
group (8/119, 6.7%)—6 in group A (6/62, 9.7%) and 2 in group B
(2/57, 3.5%). Thus, degraded HAP was more commonly seen in
group A than in group B.
Respiratory Waveform Analysis
The breath-hold grades during the precontrast and HAP are listed
in Table 2. Breath-holding difficulty was seen in 27.7% (33/119) of the
total group, 33.9% (21/62) of group A, and 21.1% (12/57) of group B
during the precontrast phase; and in 40.3% (48/119) of the total group,
43.6% (27/62) of group A, and 36.8% (21/57) of group B during the
HAP (Table 2). Group A generally showed a higher incidence of
breath-holding difficulty compared with group B, although the difference was not statistically significant (P > 0.05).
Respiratory waveform analysis revealed that the SD values during the precontrast phase and the SD value differences between the
precontrast and HAP were significantly higher in group A (P = 0.047,
P = 0.023, respectively; Table 3). In addition, during the HAP, the SD
values of group A tended to be higher than group B, although the difference was not statistically significant (P = 0.056; Table 3).
Patients With GARD
Gadoxetic acid–related dyspnea, defined as when the SD value
of the HAP was 200 greater than that of the precontrast phase without
degraded image quality in later phases (Fig. 3), was seen in 13.5%
(16/119) of the total group, 19.4% (12/62) of group A, and 7.0%
(4/57) of group B. Gadoxetic acid–related dyspnea was observed significantly more often in group A than in group B (P = 0.046, Table 2).
Among the 16 patients with GARD, overall image quality and motion
artifacts in the precontrast phase were significantly higher in group A
(all P < 0.05; Table 4). Although there was no significant difference
in image quality scores of the 2 groups during the HAP in patients with
GARD, the image quality scores were generally higher in group A than
in group B (Table 4). In the patients without GARD, the mean image
quality scores of group A were significantly higher in both phases (all
P < 0.05; Table 4).
For comparison of the image quality scores between patients
with and without GARD in the total patient group, overall image quality
and motion artifact scores in both the precontrast and HAP were significantly higher in patients with GARD than in patients without GARD
(all P < 0.05; Table 4). In other words, patients with GARD showed
poorer image quality than patients without GARD in the total group.
Among the 16 patients who experienced GARD, 5 (31.3%) demonstrated degraded HAP (4 of 12 patients in group A and 1 of 4 patients
in group B). Among the 6 group A patients who manifested degraded
HAP, 4 experienced GARD; of the 2 group B patients who showed degraded HAP, one demonstrated GARD. Thus, not all patients who experienced GARD showed degraded HAP and, vice versa, not all
patients who showed degraded HAP demonstrated GARD.
Survey Questionnaire Analysis
Among the 116 patients who reported that they held their breath
well, 71 demonstrated good breath-hold graphs (breath-hold grades 1
www.investigativeradiology.com
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
3
Investigative Radiology • Volume 00, Number 00, Month 2016
Yoo et al
FIGURE 2. Representative cases of respiratory-related graphs and gadoxetic acid–enhanced MRIs for each breath-hold grade. A, Respiratory-related
graph showing excellent breath-hold (grade 1, SD <100) of a 68-year-old woman in group B. Precontrast (B) and HAP (C) MRIs both demonstrate
excellent overall image quality (mean overall image quality scores of 1). D, Respiratory-related graph showing mild difficulty in breath-holding
(grade 2, SD = 100–200) of a 76-year-old man in group B. Precontrast (E) and HAP (F) MRIs both manifest good overall image quality (mean overall
image quality scores of 2). G, Respiratory-related graph demonstrating moderate difficulty in breath-holding (grade 3, SD = 200–300) of a 69-year-old
woman in group A. Precontrast (H) and HAP (I) MRIs show mean overall image quality scores of 2 and 3, respectively. J, Respiratory-related graph
showing severe difficulty in breath-holding (grade 4, SD >300) of a 75-year-old woman in group A. Precontrast (K) and HAP (L) MRIs manifest mean
overall image quality scores of 2.5 and 4, respectively. Figure 2 can be viewed online in color at www.investigativeradiology.com.
4
www.investigativeradiology.com
© 2016 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
Investigative Radiology • Volume 00, Number 00, Month 2016
Overcoming Degraded MR Hepatic Arterial Phase
FIGURE 3. Example of a respiratory-related graph and gadoxetic acid–enhanced MRIs of a patient with GARD. A, Respiratory-related graph showing
severe difficulty in breath-holding (grade 4, SD >300) of a 77-year-old woman with GARD in group A. The SD value difference between the precontrast
and HAP is 569.8 (greater than 200). Precontrast (B) and HAP (C) images of gadoxetic acid–enhanced MRI both demonstrate mean overall image quality
scores of 2 and 4, respectively. Portal (D) and transitional (E) phases both show good overall image quality (mean overall image quality scores of 2).
A nodular lesion manifesting low SI in the precontrast phase and continuous enhancement in the arterial, portal, and transitional phases is seen in liver
segment 4. This lesion, assumed to be a hemangioma, shows low conspicuity in the arterial phase due to poor image quality caused by motion artifact.
Therefore, poor image quality in the arterial phase may lower detection of liver lesion and affect diagnostic performance. Figure 3 can be viewed online in
color at www.investigativeradiology.com.
and 2; SD ≤ 200) while 45 showed breath-holding difficulty (breath
hold grades 3 and 4; SD > 200). Among the 45 patients who thought
they held their breath well but actually did not, 25 were in group A
and 20 were in group B. Thus, 38.8% (45/116) of the survey questionnaire replies proved to be false and thus unreliable. Furthermore, among
the 116 patients who answered that they held their breath well,
108 patients (93.1%) showed mean overall image quality scores lower
than 4, and 8 patients (6.9%) demonstrated scores of 4 (degraded image
quality). Among these 8 patients, 6 and 2 were from group A and
B, respectively (Supplement E4, Supplemental Digital Content 4,
TABLE 2. Qualitative and Quantitative Analyses of Group A and B During the Precontrast and HAP
PRE
Overall image quality†
Motion artifact†
Degraded HAP§
Grade 1 (SD <100)§
Grade 2 (SD 100–200)§
Grade 3 (SD 200–300)§
Grade 4 (SD >300)§
Breath-holding difficulty (SD >200)§
GARD§
HAP
Group A (n = 62)
Group B (n = 57)
P*
Group A (n = 62)
Group B (n = 57)
P*
2.1 (0.5)
2.1 (0.5)
NA
16/62 (25.8%)
25/62 (40.3%)
15/62 (24.2%)
6/62 (9.7%)
21/62 (33.9%)
NA
1.3 (0.4)
1.3 (0.4)
NA
22/57 (38.6%)
23/57 (40.4%)
8/57 (14.0%)
4/57 (7.0%)
12/57 (21.1%)
NA
<0.001
<0.001
NA
0.135
0.997
0.161
0.601
0.119
NA
2.4 (0.7)
2.4 (0.7)
6/62 (9.7%)
14/62 (22.6%)
21/62 (33.9%)
10/62 (16.1%)
17/62 (27.4%)
27/62 (43.6%)
12/62 (19.4%)
1.7 (0.7)
1.7 (0.8)
2/57 (3.5%)
15/57 (26.3%)
21/57 (36.8%)
12/57 (21.1%)
9/57 (15.8%)
21/57 (36.8%)
4/57 (7.0%)
<0.001‡
<0.001‡
0.421
0.635
0.681
0.489
0.125
0.456
0.046‡
*P values were calculated by using the Student t test or Fisher exact test, as appropriate.
†Data are presented as mean (SD).
‡Significant value.
§Data are the number of patients, and the value in parentheses is the percentage.
PRE indicates precontrast phase; HAP, hepatic arterial phase; NA, not applicable.
© 2016 Wolters Kluwer Health, Inc. All rights reserved.
www.investigativeradiology.com
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
5
Investigative Radiology • Volume 00, Number 00, Month 2016
Yoo et al
TABLE 3. SD Values of Group A and B
Total Group
(n = 119)
SD value of the
precontrast phase
SD value of the HAP
SD value difference
between the precontrast
and HAP (SD value of
the HAP − SD value of
the precontrast phase)
Group A
(n = 62)
Group B
(n = 57)
P*
156 (119)
176 (143) 133 (83)
0.047
213 (156)
102 (127)
239 (181) 185 (119) 0.056
139 (156) 72 (88) 0.023
Data are presented as mean (SD).
*P values were calculated by using the Student t test.
http://links.lww.com/RLI/A251 and Supplement E5, Supplemental
Digital Content 5, http://links.lww.com/RLI/A252).
DISCUSSION
There have been many efforts to overcome degraded HAP and
achieve optimal HAP image quality in gadoxetic acid–enhanced MRI
by shortening the scanning time, lowering the contrast injection rate,
using weight-based dose of contrast media, and employing optimized
protocols such as the bolus tracking technique, saline flushing, and contrast dilution.1,3,5,10,13,15–18 Among these efforts, shortening the scanning time while maintaining high resolution can be a promising
solution. Through this prospective study, we validated if minimizing
breath-hold time with short breath-hold MR techniques can be of help
in improving HAP image quality. Based on respiratory waveform analysis, group B showed lower incidences of breath-holding difficulty
(36.8% [21/57] vs 43.6% [27/62]) and GARD (7.0% [4/57] vs 19.4%
[12/62]) compared with group A. This may be explained by the fact that
the required breath-hold time for patients in group B was 5 seconds
shorter than that of patients in group A (13 vs 18 seconds). For
image-based analysis, patients in group B showed better HAP image
quality and less degraded HAP (3.5% [2/57] vs 9.7% [6/62]) than patients in group A. As patients in group B showed lower incidences of
breath-holding difficulty and GARD, this may have led to less motion
artifacts and better HAP image quality.
The incidence of degraded HAP (3.5%) and GARD (7.0%) in
group B was lower than previous studies using long breath-hold techniques (18–23 seconds), which reported incidences of 12%–18% and
10.7%–39%, respectively.6,9–13 This is probably because patients in
group B underwent an advanced parallel imaging technique, CAIPRINHA,
which allowed high resolution in a reduced acquisition time of 13 seconds.
Although the CAIPRINHA technique demonstrated lower resolution
than the conventional long breath-hold technique in this study, it
allowed a short breath-hold time without compromising resolution.
In our study, the CAIPRINHA technique showed higher resolution
compared with the MR techniques of previous studies assessing
GARD.6,9,11 Actually, the breath-hold time can be minimized to only
5 to 8 seconds with the CAIPRINHA technique. However, as it is very
difficult to sustain high resolution while minimizing the breath-hold
time, a breath-hold time of 13 seconds was considered appropriate in
maintaining a short breath-hold time while maintaining high resolution in our study. As MR techniques undergo further advances, new
techniques allowing high resolution with a breath-hold of less than
10 seconds are anticipated.
In addition, the incidence of degraded HAP and GARD in group
A was 9.7% and 19.4%, respectively, which was similar or lower than
previous studies also using long breath-hold techniques.6,9–12 This is
probably because a new 3 T MR machine using optimized protocols such
as the bolus tracking technique, saline flushing, slower injection rate
(1 mL/s), and on-label dosage (0.1 mL/kg) was used in our study, all
leading to a positive effect on breath-holding capacity and image quality.
Based on survey questionnaire responses from 116 patients who
answered that they held their breath well, 45 (38.8%) showed respiratory graphs of SD greater than 200 (breath-holding difficulty). This
suggests that some patients may have experienced dyspnea without being aware of it, which could be explained as involuntary dyspnea. However, it is unlikely that all 45 patients experienced involuntary dyspnea.
Thus, the survey questionnaire responses, which represent the subjective feeling of dyspnea, did not completely correspond to the objective
assessment of patients' respiratory conditions based on respiratory
waveform. Therefore, indirectly evaluating dyspnea by surveying the
subjective feeling of dyspnea may not be reliable. In addition, among
the 116 patients who replied that they held their breath well, 8 (6.9%)
showed image quality scores of 4 (degraded image quality). This finding implies that dyspnea may not be the only factor responsible for degraded image quality.
As 27.7% (33 patients; 21 in group A and 12 in group B) of the
total group in our study showed breath-holding difficulty in the precontrast phase, it is not surprising that 40.3% (48 patients; 27 in group
A and 21 in group B) showed breath-holding difficulty in the arterial
phase after contrast injection. Therefore, many patients, especially those
who are inpatients, heavy smokers, have chronic obstructive pulmonary
disease, or congestive heart failure, experience difficulty holding their
breath.19 As the breath-holding capacity of the majority of patients should
be taken into consideration, the breath-hold time must be shortened to
improve image quality leading to improved diagnosis regardless of
the contrast agent.
TABLE 4. Mean Image Quality Scores of Patients With and Without GARD in Precontrast and HAP
GARD (+)
GARD (−)
Total (n = 16) Group A (n = 12) Group B (n = 4)
PRE Overall image
quality
Motion artifact
HAP Overall image
quality
Motion artifact
P*
Total (n = 103) Group A (n = 50) Group B (n = 53)
P*
P†
2.2 (0.7)
2.4 (0.6)
1.5 (0.4)
<0.05
1.7 (0.6)
2.0 (0.4)
1.3 (0.4)
<0.001 <0.05
2.2 (0.7)
2.8 (0.9)
2.4 (0.5)
2.9 (0.9)
1.5 (0.4)
2.4 (0.8)
<0.05
0.288
1.7 (0.6)
1.9 (0.7)
2.0 (0.4)
2.3 (0.6)
1.3 (0.5)
1.6 (0.7)
<0.001 <0.05
<0.001 <0.001
2.8 (0.9)
2.9 (0.9)
2.4 (0.8)
0.288
1.9 (0.8)
2.3 (0.6)
1.6 (0.7)
<0.001 <0.001
Data are presented as means (SD). The numbers are averaged values between the 2 readers.
*P values were calculated between groups A and B by using the Student t test.
†P values were calculated between patients with and without GARD by using the Student t test.
6
www.investigativeradiology.com
© 2016 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
Investigative Radiology • Volume 00, Number 00, Month 2016
In previous studies, transient severe motion was used to describe
GARD based on image analysis.6,9,10 In our study, degraded HAP,
which was qualitatively evaluated based on image quality scoring,
corresponded to transient severe motion. However, GARD was quantitatively defined based on respiratory pattern analysis by comparing respiratory waveforms before and after contrast injection to evaluate the
effect of gadoxetic acid on respiration. In other words, we assessed
the difference in SD values between the precontrast and HAP to evaluate GARD. However, GARD did not completely correlate with degraded HAP image quality in our study. Objective dyspnea evaluated
by respiratory waveforms and subjective dyspnea assessed by survey
questionnaires did not completely correspond. In addition, not all objective and subjective dyspnea led to motion-related artifacts. Therefore, it
is uncertain if dyspnea can be fully evaluated by evaluating respiratory waveforms, querying subjective feelings of dyspnea, or assessing
motion-related artifacts in MRIs. Further studies are anticipated to
establish a tool for accurately assessing dyspnea.
This study had several limitations. First, the study population
was limited to a single center. Second, respiration was analyzed objectively by recording thoracoabdominal movement via a pneumatic cushion, which may not correspond to the actual respiratory state. Although
it would be ideal to record diaphragm movement, we believe that
thoracoabdominal movement is sufficient to represent the respiratory
condition to some degree. Third, the criteria or thresholds for the SD
values were determined at the discretion of the radiologist after roughly
reviewing the respiratory graphs.14 As the incidences of breath-holding
difficulty and GARD rely greatly on this threshold value, a more objective criterion and further studies on this matter are required. Fourth, only
the respiratory waveforms of the precontrast and HAP were evaluated in
this study. As the respiratory waveforms of the portal and transitional
phases were not evaluated, it was unclear if the dyspnea was transient
or retained in later phases. Therefore, we evaluated the image quality
of the portal and transitional phases to confirm that image quality degradation was transient and not retained in the phases following the HAP.
Fifth, the impact of image quality on liver lesion detection and assessment of the 2 groups was not evaluated. Further within-patient studies
comparing the diagnostic performance of the 2 groups are anticipated.
In summary, respiratory waveform analysis revealed incidences
of breath-holding difficulty of 43.6% and 36.8% in group A and B, respectively. Gadoxetic acid–related dyspnea during the HAP was seen in
19.4% of group A and 7.0% of group B. Image-based analysis demonstrated degraded HAP in 9.7% of group A and 3.5% of group B. Therefore, the respiratory waveform and image quality analyses in this study
indicated that the short breath-hold MR technique, CAIPIRINHA,
showed better HAP image quality with less degraded HAP and a lower
incidence of breath-holding difficulty and GARD than the conventional long breath-hold technique. The difference in breath-hold time between groups A and B was only 5 seconds (18 vs 13 seconds). Although
5 seconds may seem like a short time, it significantly affected HAP image
quality during gadoxetic acid–enhanced MRI. This may be the “magic of
5 seconds.” Therefore, we believe the short breath-hold MR technique,
CAIPIRINHA, can be the first step to overcoming degraded HAP in
gadoxetic acid–enhanced liver MRI.
ACKNOWLEDGMENTS
The authors thank Yun Mi Seo, MS, for conducting the respiratory graph.
© 2016 Wolters Kluwer Health, Inc. All rights reserved.
Overcoming Degraded MR Hepatic Arterial Phase
REFERENCES
1. Park YS, Lee CH, Kim IS, et al. Usefulness of controlled aliasing in parallel
imaging results in higher acceleration in gadoxetic acid-enhanced liver magnetic resonance imaging to clarify the hepatic arterial phase. Invest Radiol.
2014;49:183–188.
2. Goodwin MD, Dobson JE, Sirlin CB, et al. Diagnostic challenges and pitfalls in
MR imaging with hepatocyte-specific contrast agents. Radiographics. 2011;31:
1547–1568.
3. Ringe KI, Husarik DB, Sirlin CB, et al. Gadoxetate disodium-enhanced MRI of
the liver: part 1, protocol optimization and lesion appearance in the noncirrhotic
liver. AJR Am J Roentgenol. 2010;195:13–28.
4. Willatt JM, Hussain HK, Adusumilli S, et al. MR imaging of hepatocellular carcinoma in the cirrhotic liver: challenges and controversies. Radiology. 2008;247:
311–330.
5. Cruite I, Schroeder M, Merkle EM, et al. Gadoxetate disodium-enhanced MRI of
the liver: part 2, protocol optimization and lesion appearance in the cirrhotic liver.
AJR Am J Roentgenol. 2010;195:29–41.
6. Davenport MS, Viglianti BL, Al-Hawary MM, et al. Comparison of acute transient dyspnea after intravenous administration of gadoxetate disodium and
gadobenate dimeglumine: effect on arterial phase image quality. Radiology. 2013;
266:452–461.
7. Seale MK, Catalano OA, Saini S, et al. Hepatobiliary-specific MR contrast
agents: role in imaging the liver and biliary tree. Radiographics. 2009;29:
1725–1748.
8. Frydrychowicz A, Nagle SK, D'Souza SL, et al. Optimized high-resolution
contrast-enhanced hepatobiliary imaging at 3 tesla: a cross-over comparison of
gadobenate dimeglumine and gadoxetic acid. J Magn Reson Imaging. 2011;34:
585–594.
9. Bashir MR, Castelli P, Davenport MS , et al. Respiratory motion artifact affecting
hepatic arterial phase MR imaging with gadoxetate disodium is more common in
patients with a prior episode of arterial phase motion associated with gadoxetate
disodium. Radiology. 2015;274:141–148.
10. Pietryga JA, Burke LM, Marin D, et al. Respiratory motion artifact affecting hepatic arterial phase imaging with gadoxetate disodium: examination
recovery with a multiple arterial phase acquisition. Radiology. 2014;271:
426–434.
11. Davenport MS, Caoili EM, Kaza RK, et al. Matched within-patient cohort study
of transient arterial phase respiratory motion-related artifact in MR imaging
of the liver: gadoxetate disodium versus gadobenate dimeglumine. Radiology.
2014;272:123–131.
12. Davenport MS, Bashir MR, Pietryga JA, et al. Dose-toxicity relationship of
gadoxetate disodium and transient severe respiratory motion artifact. AJR Am J
Roentgenol. 2014;203:796–802.
13. Hayashi T, Saitoh S, Tsuji Y, et al. Influence of gadoxetate disodium on oxygen
saturation and heart rate during dynamic contrast-enhanced MR imaging. Radiology. 2015;276:756–765.
14. Park YS, Lee CH, Yoo JL , et al. Hepatic arterial phase in gadoxetic acidenhanced liver magnetic resonance imaging: analysis of respiratory patterns and
their effect on image quality. Invest Radiol. 2016;51:127–133.
15. Chandarana H, Feng L, Ream J, et al. Respiratory motion-resolved compressed
sensing reconstruction of free-breathing radial acquisition for dynamic liver magnetic resonance imaging. Invest Radiol. 2015;50:749–756.
16. Haradome H, Grazioli L, Tsunoo M, et al. Can MR fluoroscopic triggering technique and slow rate injection provide appropriate arterial phase images with reducing artifacts on gadoxetic acid-DTPA (Gd-EOB-DTPA)-enhanced hepatic MR
imaging? J Magn Reson Imaging. 2010;32:334–340.
17. Schmid-Tannwald C, Herrmann K, Oto A , et al. Optimization of the dynamic,
Gd-EOB-DTPA-enhanced MRI of the liver: the effect of the injection rate. Acta
Radiol. 2012;53:961–965.
18. Kazmierczak PM, Theisen D, Thierfelder KM, et al. Improved detection of
hypervascular liver lesions with CAIPIRINHA-Dixon-TWIST-volume-interpolated
breath-hold examination. Invest Radiol. 2015;50:153–160.
19. Gay SB, Sistrom CL, Holder CA, et al. Breath-holding capability of adults.
Implications for spiral computed tomography, fast-acquisition magnetic resonance imaging, and angiography. Invest Radiol. 1994;29:848–851.
www.investigativeradiology.com
Copyright © 2016 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
This paper can be cited using the date of access and the unique DOI number which can be found in the footnotes.
7