DOI:10.2214/AJR.05.0852
AJR 2006; 187:618-622
© American Roentgen Ray Society
Effects of High-Resolution CT of the Lung Using Partial Versus Full Reconstruction on Motion Artifacts and Image Noise
Hong Il Ha1,
Hyun Woo Goo1,
Joon Beom Seo1,
Jae-Woo Song1 and
Jin Seong Lee1
1 All authors: Department of Radiology, Asan Medical Center, University of Ulsan
College of Medicine, 388-1 Poongnap-2 dong, Songpa-gu, Seoul, South Korea
138-736.
Received May 19, 2005;
accepted after revision July 12, 2005.
Address correspondence to H. W. Goo
(hwgoo{at}amc.seoul.kr).
Abstract
OBJECTIVE. The purpose of our study was to evaluate the effects of
0.3-second high-resolution CT (HRCT) of the lung using partial reconstruction
on cardiac motion artifacts and image noise.
SUBJECTS AND METHODS. Thirty-seven pairs of 0.3-second (partial
reconstruction) and 0.75-second (full reconstruction) HRCT images were
obtained for the lower lung zone during full-inspiration breath-holding.
Imaging parameters other than temporal resolution were identical for each
patient. Two radiologists visually graded motion artifacts of the cardiac
border, bronchi, pulmonary vessels, and fissure in the left lung on a 4-point
scale (with 4 indicating no artifacts). The maximum width of motion along the
left cardiac border and the area percentage of motion artifacts in the left
lung were calculated. Image noise in the air and lung was also determined.
Cardiac motion artifacts and image noises were compared between the two sets
of CT images.
RESULTS. Visual grades for the cardiac border (4 ± 0),
bronchi (3.8 ± 0.7), pulmonary vessels (3.6 ± 0.8), and fissure
(3.9 ± 0.5) were higher for 0.3-second images than for 0.75-second
images (1.7 ± 0.7, 2.0 ± 1.0, 1.6 ± 0.7, and 2.4 ±
0.9, respectively) (p < 0.001). The maximum width of motion along
the left cardiac border (0.1 ± 0.5 mm) and the area percentage of
motion artifacts in the left lung (6.7% ± 18.4%) were smaller for
0.3-second images than for 0.75-second images (4.5 ± 1.7 mm and 36.2%
± 20.9%, respectively) (p < 0.001). Image noises in the air
(38.0 ± 9.2) and the lung (86.0 ± 23.1) were greater for
0.3-second images than for 0.75-second images (35.6 ± 9.6 and 76.0
± 20.3, respectively) (p < 0.01).
CONCLUSION. Compared with 0.75-second HRCT using full
reconstruction, 0.3-second HRCT using partial reconstruction substantially
reduces cardiac motion artifacts in the lung at the expense of increasing
image noise.
Keywords: CT technique • high-resolution CT • lung
Introduction
High-resolution CT (HRCT) of the lung is the accepted diagnostic method for
the detection and characterization of various pulmonary parenchymal
abnormalities involving the airways, air space, and interstitium
[1-4].
HRCT image quality is substantially affected by respiratory motion artifacts,
cardiac motion artifacts, and radiation dose. Respiratory motion artifacts can
be virtually eliminated if patients hold their breath during the scan, and
better breath-holding may be achieved by hyperventilation and administration
of oxygen before scanning. Regardless of breath-holding, cardiac motion
artifacts can still affect lung images, particularly in the paracardiac
regions, and may lead to misinterpretation (e.g., bronchiectasis)
[2,
5-7].
Cardiac motion artifacts can be reduced by the use of shorter gantry rotation
times, prospective ECG triggering, retrospective ECG gating, and partial
reconstruction
[8-11].
Although associated with an increase in image noise, low-dose HRCT has been
reported to provide diagnostic-quality images
[12,
13]. Modern CT machines are
able to achieve shorter gantry rotation times up to 0.33 seconds, and as a
result half-temporal resolution (approximately 0.22 seconds) of the HRCT scan
can be obtained using partial reconstruction. The purpose of this prospective
study was to evaluate 0.3-second HRCT of the lung using partial reconstruction
in terms of cardiac motion artifacts and image noise.
Subjects and Methods
Between February 2004 and March 2004, 43 consecutive patients referred for
HRCT were enrolled in this study. The institutional review board approved the
study and informed consent was obtained from all patients. All patients
underwent HRCT during single or clustered full-inspiration breath-holding
using a 16-MDCT (Somatom Sensation 16, Siemens Medical Solutions) scanner that
was capable of partial (0.3-, 0.36-, or 0.54-second) or full (0.42-, 0.5-, or
0.75-second) reconstruction with variable gantry rotation times. HRCT images
with partial reconstruction used data from 240° partial rotation rather
than 360° full rotation. We chose 0.3-second HRCT images using partial
reconstruction from 0.42-second gantry rotation time and 0.75-second HRCT
images using data from 360° full rotation because we believe that
0.3-second gantry rotation time may represent the shortest temporal resolution
for the modern CT system and 0.75-second gantry rotation time may represent
the usual or average temporal resolution for the most widely used CT
system.
Imaging parameters other than temporal resolution were identical for each
patient. Section thickness was 1 mm for all patients, with fixed intervals of
10 mm used for adults, and flexible intervals according to body size, with 10
slices covering the whole lung used for children. Acquisition parameters of 85
kVp and 140 effective mAs were used for adults; weight-based parameters were
used for children (Table 1).
Different kernels (B70f and B50s) were used for adults and children, and an
appropriate field of view was used according to body size.
After HRCT was performed with a 0.3-second protocol, one 0.75-second HRCT
image was added, obtained in the lower lung zone at one of the slice positions
of the 0.3-second HRCT. Six patients were excluded because of significant
respiratory misregistration between paired HRCT images by means of a visual
assessment, which was thought substantially to affect accurate comparison
between them. Therefore, 37 patients (22 males and 15 females; age range, 7-81
years; mean age, 44.5 years) were enrolled in this study. Patient diagnoses
included primary or secondary malignancy (n = 12), bronchiectasis
(n = 5), benign pulmonary nodule (n = 4), lymphadenopathy
(n = 3), cardiac disease (n = 3), and others (n =
10).
We evaluated only the left lower lung zone because it is most prone to
cardiac motion artifacts [14],
and this approach also minimized patients' radiation exposure. For analysis,
technical parameters regarding HRCT images were hidden on a PACS workstation.
Moreover, 74 HRCT images were presented in random order. For subjective
analysis, two radiologists in consensus visually graded the motion artifacts
of the cardiac border, pulmonary vessels, bronchi, and fissure in the left
lung of HRCT images using a 4-point scale
(Table 2). For objective
analysis, the maximum width of motion along the left cardiac border and the
area percentage of motion artifacts in the left lung were measured on the PACS
by one radiologist (Fig. 1). We
determined the area containing motion artifacts in the left lung by drawing
the border between the areas with and without motion artifacts, usually at the
paracardiac area of the left lung. Objective analysis was performed 1 week
after subjective analysis.
View this table:
[in this window]
[in a new window]
|
TABLE 2: Visual Grades of Cardiac Motion Artifacts in Lung (Cardiac Border,
Bronchi, Pulmonary Vessels, and Fissure) on HRCT Images
|
|
View larger version (81K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1 —34-year-old man with benign pulmonary nodule. High-resolution
CT image of left lung exemplifies measurement of maximum width
(arrows) of cardiac motion artifact and area percentage of motion
artifacts in left lung. Artifact area (dotted line) in left lung was
bordered by outermost points of paracardiac area showing cardiac motion
artifacts, and its percentage to whole area of left lung at scanned level was
calculated.
|
|
Image noise (SD of CT densitometry) was measured in the air anterior to the
upper chest wall and in the lung (the left lower lobe) within a rectangular
region of interest (130-220 mm2). The location and size of the
region of interest were the same for each patient. We placed the region of
interest in the lung with careful attention to exclude larger pulmonary
vessels. Image noises were measured twice and averaged.
Statistical analyses were performed using SPSS version 10.1 (Statistical
Package for the Social Sciences) for Windows (Microsoft). A two-tailed
Student's t test was used for comparing continuous variables, and
Wilcoxon's signed rank test was used for comparing ordinal variables. A
p value of less than 0.05 was considered to indicate a significant
difference.
Results
We found that visual grades of the cardiac border (4 ± 0), bronchi
(3.8 ± 0.7), pulmonary vessels (3.6 ± 0.8), and fissure (3.9
± 0.5) for 0.3-second HRCT images were higher than those for
0.75-second HRCT images (1.7 ± 0.7, 2.0 ± 1.0, 1.6 ± 0.7,
2.4 ± 0.9, respectively) (p < 0.001) (Figs.
2,
3A, and
3B). Furthermore, the total
visual grade of all evaluated anatomic structures was higher for 0.3-second
HRCT images (15.4 ± 1.2) than for 0.75-second HRCT images (7.6 ±
2.5) (p < 0.001). The maximum width of cardiac motion along the
left cardiac border and the area percentage of the portion affected by motion
artifacts in the left lung were smaller for 0.3-second HRCT images (0.1
± 0.5 mm and 6.7% ± 18.4%, respectively) than for 0.75-second
images (4.5 ± 1.7 mm and 36.2% ± 20.9%, respectively)
(p < 0.001) (Figs.
4 and
5). Image noises for the air
and the lung parenchyma were greater for 0.3-second HRCT images (38.0 ±
9.2 and 86.0 ± 23.1, respectively) than for 0.75-second images (35.6
± 9.6 and 76.0 ± 20.3, respectively) (p < 0.01)
(Fig. 6), which translated to
mean increases of 6.7% for the air and 10.0% for the lung.
View larger version (11K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2 —Visual grading of cardiac motion artifacts in cardiac border,
bronchi, vessels, and fissure on 0.3-second (gray bars) and
0.75-second (black bars) high-resolution CT (HRCT) images. Difference
in visual grade for each anatomic structure between 0.3-second and 0.75-second
HRCT images was significant (p < 0.001). Difference in total
visual grade of all evaluated anatomic structures between 0.3-second images
(15.4 ± 1.2) and 0.75-second images (7.6 ± 2.5) was significant
(p < 0.001).
|
|
View larger version (96K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —76-year-old man with lung cancer. 0.75-second high-resolution
CT image shows motion artifacts along cardiac border (> 4 mm in maximum
width) and doubling (arrows) of larger bronchi, larger vessels, and
fissure.
|
|
View larger version (7K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —Maximum width of cardiac motion artifact along left cardiac
border in 0.3-second and 0.75-second high-resolution CT (HRCT) images.
Difference between 0.3-second images (0.1 ± 0.5 mm) and 0.75-second
images (4.5 ± 1.7 mm) was significant (p < 0.001).
|
|
View larger version (7K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5 —Area percentage of motion artifact in left lung in 0.3-second
and 0.75-second high-resolution CT (HRCT) images. Difference between
0.3-second images (6.7% ± 18.4%) and 0.75-second images (36.2% ±
20.9%) was significant (p < 0.001).
|
|
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6 —Image noises (standard deviation of CT densitometry) in air
and lung in 0.3-second (gray bars) and 0.75-second (black
bars) high-resolution CT (HRCT) images. Image noises in air (38.0
± 9.2) and lung parenchyma (86.0 ± 23.1) for 0.3-second HRCT
images were greater than those on 0.75-second images (35.6 ± 9.6 and
76.0 ± 20.3, respectively) (p < 0.01).
|
|
Discussion
Modern CT systems capable of shorter gantry rotation times would be
expected to produce fewer motion-related artifacts on HRCT images. The
temporal resolution of HRCT scans can be further enhanced by means of partial
rotation reconstruction. In the present study, we found that 0.3-second HRCT
using partial reconstruction resulted in fewer cardiac motion artifacts in the
lung compared with 0.75-second HRCT using full reconstruction. Anatomic
details in paracardiac areas were clearly depicted on 0.3-second HRCT images.
This is likely to result in greater diagnostic accuracy for lung abnormalities
because some cardiac motion artifacts in HRCT of the lung mimic pathologic
conditions and thus constitute a potential cause of misdiagnosis
[2,
5-7].
Other investigators reported that an acquisition time of less than 19.1
milliseconds would be necessary for complete elimination of all motion
artifacts [15]. In the current
study, we evaluated only cardiac motion artifacts on breath-holding HRCT
images of the lung and found that a temporal resolution of 0.3 seconds
eliminated virtually all cardiac motion artifacts. There may be two possible
explanations for the remaining motion artifacts in 0.3-second HRCT scans.
First, a temporal resolution of 0.3 seconds may not be sufficient for complete
elimination of cardiac motion artifacts. Second, the lower lung zone may be
affected by small diaphragmatic motions even during breath-holding.
Partial rotation reconstruction has been considered an effective way to
reduce motion-related artifacts. We found that HRCT using partial
reconstruction was successful even in children who could hold their breath for
only 3-5 seconds. Some authors reported that 0.6-second breath-hold HRCT using
multiple-segment partial reconstruction reduced cardiac motion artifacts
without increasing patient dose
[11]. In addition, other
researchers found that 0.6-second CT images using partial reconstruction
provided better image quality by reducing motion artifacts than 1.0-second CT
images using full reconstruction in patients being treated with mechanical
pulmonary ventilation who could not hold their breath
[16].
We speculate that HRCT using partial reconstruction also significantly
reduces respiratory and cardiac motion artifacts in free-breathing young
children without increasing radiation dose. Because respiratory motion appears
to affect HRCT image quality more than cardiac motion, respiration-triggered
scans may be more beneficial for free-breathing young children than
ECG-triggered scans. Although respiration-triggered CT has been attempted in a
few studies [17], it is not
routinely used. However, a further study should be performed regarding
clinical applications of respiration-triggered HRCT because it is a promising
and effective method for reducing respiratory motion on HRCT images in
uncooperative patients.
In addition, we think that we may extend our results to helical CT because
almost all cardiac motion artifacts in the lung may be gone at a 0.3-second
temporal resolution regardless of 0.3-second partial rotation or 0.3-second
full rotation. Therefore, we believe that similar effects on cardiac motions
may be found on helical chest CT with a 0.3-second gantry rotation time.
ECG-triggered scanning is one way to reduce cardiac motion artifacts, and
acquisition windows should be set to the quietest period of the cardiac cycle,
usually mid-diastole [10].
Some authors found that ECG triggering reduced cardiac motion artifacts,
particularly in the left lower lung, compared with nongated HRCT scans of the
lung with the same temporal resolution of 0.5 seconds
[14]. Further study is
necessary to determine whether ECG triggering still has an additional benefit
in reducing cardiac motion artifacts in the current faster CT systems. In
contrast, ECG-gated scanning is not recommended for HRCT because it delivers a
large amount of radiation
[18].
Increased image noise is a trade-off of increased temporal resolution. In
the present study, image noise on 0.3-second HRCT images was greater than that
on 0.75-second HRCT images by 6.7% in the air and 10.0% in the lung. However,
increased image noise on low-dose HRCT images is not likely to compromise
diagnostic information in the lung except in cases of low-contrast lesions,
such as subtle ground-glass opacity and mild emphysema
[13,
19,
20].
In conclusion, compared with 0.75-second HRCT using full reconstruction,
0.3-second HRCT using partial reconstruction reduces cardiac motion artifacts
in the lung at the expense of increasing image noise.
References
- Gutierrez FR, Woodard PK, Fleishman MJ, Semenkovich JW, Anderson
DC. Thorax: techniques and normal anatomy. In: Lee JKT, Sagel SS, Stanley RJ,
Heiken JP, eds. Computed body tomography with MRI correlation,
3rd ed. Philadelphia, PA: Lippincott-Raven, 1998:183
-260
- Naidich DP, MaCauley DI, Khouri NF, Stitik FP, Siegelman SS. CT of
bronchiectasis. J Comput Assist Tomogr1982; 6:437
-444[Medline]
- Naidich DP, Zerhouni EA, Hutchins GM, Genieser NB, McCauley DI,
Siegelman SS. CT of the pulmonary parenchyma. Part 1. Distal air-space
disease. J Thorac Imaging 1985;1
: 39-53[Medline]
- Zerhouni EA, Naidich DP, Stitik FP, Khounri NF, Siegelman SS.
Computed tomography of the pulmonary parenchyma. Part II. Interstitial
disease. J Thorac Imaging 1985;1
: 54-64[Medline]
- Tarver RD, Conces DJ Jr, Godwin JD. Motion artifacts on CT simulate
bronchiectasis. AJR 1988;151
: 1117-1119[Free Full Text]
- Mayo JR, Müller NL, Henkelman RM. The double-fissure sign: a
motion artifact on thin-section CT scans. Radiology1987; 165:580
-581[Abstract/Free Full Text]
- Kuhns LR, Borlaza G. The "twinkling star" sign: an aid
in differentiating pulmonary vessels from pulmonary nodules on computed
tomograms. Radiology 1980;135
: 763-764[Abstract/Free Full Text]
- Rubin GD, Leung AN, Robertson VJ, Stark P. Thoracic helical CT:
influence of subsecond gantry rotation on image quality.
Radiology 1998;208
: 771-776[Abstract/Free Full Text]
- Hofmann LK, Zou KH, Costello P, Schoepf UJ.
Electrocardiographically gated 16-section CT of the thorax: cardiac motion
suppression. Radiology 2004;233
: 927-933[Abstract/Free Full Text]
- Boehm T, Willmann JK, Hilfiker PR, et al. Thin-section CT of the
lung: does ECG triggering influence diagnosis?
Radiology 2003;229
: 483-491[Abstract/Free Full Text]
- Arac M, Oner AY, Celik H, Akpek S, Isik S. Lung at thin-section CT:
influence of multiple-segment reconstruction on image quality.
Radiology 2003;229
: 195-199[Abstract/Free Full Text]
- Zwirewich CV, Mayo JR, Müller NL. Low-dose high-resolution CT
of lung parenchyma. Radiology 1991;108
: 413-417
- Lucaya J, Piqueras J, García-Peña P, Enríquez
G, García-Macías M, Sotil J. Low-dose high-resolution CT of the
chest in children and young adults: dose, cooperation, artifact incidence, and
image quality. AJR 2000;175
: 985-992[Abstract/Free Full Text]
- Montaudon M, Berger P, Blachere H, De Boucaud L, Latrabe V, Laurent
F. Thin-section CT of the lung: influence of 0.5-s gantry rotation and ECG
triggering on image quality. Eur Radiol2001; 11:1681
-1687[CrossRef][Medline]
- Ritchie CJ, Godwin JD, Crawford CR, Stanford W, Anno H, Kim Y.
Minimum scan speeds for suppression of motion artifacts in CT.
Radiology 1992;185
: 37-42[Abstract/Free Full Text]
- Posniak HV, Olson MC, Demos TC, Pierce KL, Kalbhen CL. CT of the
chest and abdomen in patients on mechanical pulmonary ventilators: quality of
images made at 0.6 versus 1.0 sec. AJR1994; 164:1073
-1077
- Ritchie CJ, Hsieh J, Gard MF, Godwin JD, Kim Y, Crawford CR.
Predictive respiratory gating: a new method to reduce motion artifacts on CT
scans. Radiology 1994;190
: 847-852[Abstract/Free Full Text]
- Trabold T, Buchgeister M, Kuttner A, et al. Estimation of radiation
exposure in 16-detector row CT of the heart with retrospective ECG-gating.
Rofo 2003; 175:1051
-1055[Medline]
- Prasad SR, Wittram C, Shepard JA, McLoud T, Rhea J. Standard-dose
and 50%-reduced-dose chest CT: comparing the effect on image quality.
AJR 2002; 179:461
-465[Abstract/Free Full Text]
- Lee KS, Primack SL, Staples CA, Mayo JR, Aldrich JE, Müller
NL. Chronic infiltrative lung disease: comparison of diagnostic accuracies of
radiography and low- and conventional-dose thin-section CT.
Radiology 1994;191
: 669-673[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
