DOI:10.2214/AJR.05.1535
AJR 2007; 188:122-129
© American Roentgen Ray Society
Polyp Measurement with CT Colonography: Multiple-Reader, Multiple-Workstation Comparison
Brett M. Young1,
J. G. Fletcher2,
Scott R. Paulsen1,
Fargol Booya2,
C. Daniel Johnson2,
Kristina T. Johnson2,3,
Zackary Melton2,
Drew Rodysill2,4 and
Jay Mandrekar5
1 Mayo Clinic College of Medicine, Rochester, MN 55905.
2 Department of Radiology, Mayo Clinic Rochester, Mayo East 2-B, 200 First St.
SW, Rochester, MN 55905.
3 Present address: St. Olaf College, Northfield, MN.
4 Present address: Northwestern University, Evanston, IL.
5 Division of Biostatistics, Mayo Clinic Rochester, Rochester, MN 55905.
Received August 30, 2005;
accepted after revision December 7, 2005.
Presented at the 2005 meeting of the Society of Gastrointestinal
Radiologists, San Antonio, Texas. Winner of the Alexander R. Margulis First
Time Paper Presenter Award.
J. G. Fletcher receives grant support from Siemens Medical Solutions and
has an educational license with GE Healthcare. C. D. Johnson has a software
license with GE Healthcare. J. G. Fletcher and C. D. Johnson teach in a CME
course using Vitrea workstations. These authors with conflicts of interest
consequently did not participate in data acquisition or data analysis as part
of the study discussed in this article.
Address correspondence to J. G. Fletcher
(fletcher.joel{at}mayo.edu).
Abstract
OBJECTIVE. The risk of invasive colorectal cancer in colorectal
polyps correlates with lesion size. Our purpose was to define the most
accurate methods for measuring polyp size at CT colonography (CTC) using three
models of workstations and multiple observers.
MATERIALS AND METHODS. Six reviewers measured 24 unique polyps of
known size (5, 7, 10, and 12 mm), shape (sessile, flat, and pedunculated), and
location (straight or curved bowel segment) using CTC data sets obtained at
two doses (5 mAs and 65 mAs) and a previously described colonic phantom model.
Reviewers measured the largest diameter of polyps on three proprietary
workstations. Each polyp was measured with lung and soft-tissue windows on
axial, 2D multiplanar reconstruction (MPR), and 3D images.
RESULTS. There were significant differences among measurements
obtained at various settings within each workstation (p < 0.0001).
Measurements on 2D images were more accurate with lung window than with
soft-tissue window settings (p < 0.0001). For the 65-mAs data set,
the most accurate measurements were obtained in analysis of axial images with
lung window, 2D MPR images with lung window, and 3D tissue cube images for
Wizard, Advantage, and Vitrea workstations, respectively, without significant
differences in accuracy among techniques (0.11 < p < 0.59). The
mean absolute error values for these optimal settings were 0.48 mm, 0.61 mm,
and 0.76 mm, respectively, for the three workstations. Within the
ultralow-dose 5-mAs data set the best methods for Wizard, Advantage, and
Vitrea were axial with lung window, 2D MPR with lung window, and 2D MPR with
lung window, respectively. Use of nearly all measurement methods, except for
the Vitrea 3D tissue cube and the Wizard 2D MPR with lung window, resulted in
undermeasurement of the true size of the polyps.
CONCLUSION. Use of CTC computer workstations facilitates accurate
polyp measurement. For routine CTC examinations, polyps should be measured
with lung window settings on 2D axial or MPR images (Wizard and Advantage) or
3D images (Vitrea). When these optimal methods are used, these three
commercial workstations do not differ significantly in acquisition of accurate
polyp measurements at routine dose settings.
Keywords: cancer • colon • CT colonography
Introduction
Computed tomography colonography (CTC) has been shown to be an effective
structural examination of the colon after incomplete endoscopy and compares
favorably with established colorectal cancer screening alternatives
[1-4].
When a polyp is found on CTC, referral for endoscopic polypectomy is warranted
for lesions larger than 1 cm in diameter, for which the risk of malignancy
exceeds 4% and the risk of high-grade dysplasia is greater than 20%. For
polyps measuring 5-9 mm, however, in which the risk of malignancy is 1% and
the risk of dysplasia is considerably less at 2-7%, referring clinicians do
not always uniformly recommend endoscopy, often balancing the risks of
endoscopy and polypectomy with those of other comorbid conditions (e.g.,
advanced age, anticoagulation, and previous difficulty with sedation)
[5-8].
In this clinical setting, accurate polyp measurement is essential in helping
referring clinicians gauge the risks and benefits of recommending endoscopic
polypectomy once a polyp smaller than 1 cm is found on CTC.
Accurate differentiation of polyps measuring 5-10 mm can be critical for
treatment choice at CTC, because accurate measurement affects triage of
patients to follow-up imaging versus immediate optical colonoscopy
[9]. Our purpose was to define
the most accurate and reproducible method of measuring polyps on CTC using
multiple reviewers and computer workstations.
Materials and Methods
Six reviewers measured polyps of known size using a phantom colon model
[10]. The phantom was
constructed of glass and designed to model a colon 5 cm in diameter. Polyps
were purposefully placed along curved and straight segments and on haustral
folds (Figs. 1A and
1B). The air-filled phantom was
submerged in a 28-cm-diameter water bath (with attenuation of the water bath
simulating the body cavity before scanning.
The phantom contained polyps mechanically milled to precise sizes and
shapes from plastic polymer (Solidwater, Gammex RMI). At our acquisition
parameters, attenuation of a solid block of polymer was 7.5 ± 14.6 H at
65 mAs and 7.1 ± 60 H at 5 mAs. Twenty-four polyps of four known size
groups (5, 7, 10, and 12 mm) and of three uniform shapes (pedunculated,
sessile, and flat) were used for measurement comparisons. Pedunculated polyps
were defined as having a stalk connecting to the colon wall and of less
circumference than the polyp head (head diameter equal to stalk height in our
phantom); sessile polyps as having a base diameter equal to polyp height; and
flat polyps as having a base diameter twice polyp height. Polyps of each size
and shape were selected within curved and straight segments of the phantom.
The phantom was scanned with an 8-MDCT scanner (LightSpeed Ultra, GE
Healthcare) at doses of 5 mAs and 65 mAs (2.5-mm slice thickness, 1.25-mm
reconstruction interval, pitch of 1.35, 27 mm/s table speed, 120 kVp,
0.5-second rotation time, 10-mA and 130-mA tube currents).
Three workstations were studied: Wizard (Syngo colonography, Siemens
Medical Solutions), Advantage Windows 4.2 (3.0.58F Voxtool colonography, GE
Healthcare), and Vitrea 2 (colonography version 3.5, Vital Images). Each of
the six reviewers measured polyps with two of the three workstations
(reviewers 1 and 2 used Advantage and Wizard; reviewers 3 and 4 used Vitrea
and Advantage; and reviewers 5 and 6 used Wizard and Vitrea). Five of the six
reviewers had completed a previously described CTC training course using a
laptop computer module consisting of 50 teaching cases and 50 full CTC data
sets [11]. The sixth reviewer
had previous training with more than 50 clinical cases. Each reviewer
practiced with the assigned workstation before taking measurements. A third
investigator, the same in all cases, observed and assisted the reviewers.
Reviewers independently measured 24 polyps using three methods: 2D axial,
2D multiplanar reconstruction (MPR), and 3D volume renderings.
Three-dimensional measurements were not taken using the Wizard workstation,
which does not support measurements from its 3D perspective, volume-rendered
images. Axial and MPR measurements were obtained at default lung (Advantage,
1,500/-600; Vitrea, 1,500/-200; Wizard, 1,500/-400) and soft-tissue
(Advantage, 400/20; Vitrea, 400/10; Wizard, 400/40) window settings. All
polyps were randomized and labeled before measurement to minimize the chance
that a reviewer would measure the wrong polyp. Polyps from each data set were
measured five times by each reviewer.
Axial Measurement Method
Selecting from a series of axial images, reviewers determined which image
showed the largest dimension of a polyp. The largest diameter was then
measured with the linear tool of the workstation. Pedicles of pedunculated
polyps were not included as part of the measurement. Measurements were
obtained with both lung and soft-tissue windows for the 2D views (axial and
MPR) (Figs. 2A,
2B,
2C,
2D,
2E,
2F,
2G,
2H, and
2I).
View larger version (75K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2F —Colon phantom. Two-dimensional axial CT images show method
for determining polyp size by measurement of largest diameter of polyp with
linear tool. Contiguous images at soft-tissue window.
|
|
View larger version (76K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2G —Colon phantom. Two-dimensional axial CT images show method
for determining polyp size by measurement of largest diameter of polyp with
linear tool. Contiguous images at soft-tissue window.
|
|
View larger version (79K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2H —Colon phantom. Two-dimensional axial CT images show method
for determining polyp size by measurement of largest diameter of polyp with
linear tool. Contiguous images at soft-tissue window.
|
|
View larger version (70K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2I —Colon phantom. Two-dimensional axial CT images show method
for determining polyp size by measurement of largest diameter of polyp with
linear tool. Contiguous images at soft-tissue window.
|
|
MPR Measurement Method
In a manner similar to that used for axial measurement, reviewers selected
from sagittal, coronal, and oblique series of polyps, measuring from the image
displaying the largest diameter (Figs.
3A,
3B,
3C,
3D,
3E,
3F,
3G,
3H,
3I, and
3J). This dimension was
measured with the linear tool of the workstation.
3D Measurement Method
Measurements from 3D images were obtained with a 3D tissue cube (Vital
Images) and endoluminal perspective, volume-rendered images (GE Healthcare).
Reviewers viewed all angles of a polyp in 3D, changing the viewing angle to
allow the virtual lighting to expose the true edge of a polyp for selection of
the image best depicting the largest dimension (Figs.
3K,
3L,
3M,
3N,
3O, and
3P). Reviewers obtained the
measurements with the 3D linear tool of the workstations. Stalks were not
included. Reviewers were instructed to use judgment in cursor placement while
taking 3D measurements, because even slight misplacement of the cursor from
the edge of a polyp to the far wall of the phantom might have led to absurdly
large measurements.
Statistical Analysis
Errors in measurements were calculated with the difference between the
measured and the actual size of a polyp, the negative or positive value
assigned to these differences being retained. Graphical analysis was performed
to show the error of each method per workstation for the 65-mAs and 5-mAs data
sets. Further statistical analysis was based on absolute differences rather
than simple differences, because positive and negative error values might have
canceled each other. Within each dose level and within each setting of a
workstation, we identified the settings that might be considered the best
choices. Analysis of variance for repeated measures was performed to compare
settings (analysis of variance with multiple settings as a repeat factor). The
results for the settings identified within a workstation were then compared by
paired Student's t test. This procedure was performed for each dose
level separately. Agreement between reviewers was estimated by intraclass
correlation and calculated with the combined data from the 65-mAs and 5-mAs
data sets.
Results
Error (measured size minus true size) for each method per workstation for
the 65-mAs and 5-mAs data sets is shown in Figures
4 and
5. Results from one-factor
repeated-measures analysis of variance in the case of each of the three
workstations indicated significant differences among measurements at various
settings within each machine (p < 0.0001). For all workstations
and both dose settings, axial and MPR measurements obtained at lung window
settings were more accurate than measurements obtained at soft-tissue-window
settings (p < 0.0001).
View larger version (16K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —Box plot shows error values (measured size minus true size)
for 65-mAs data set. Box represents 50% of values. Horizontal line = median
value, vertical line = range of values, GE = GE Healthcare Advantage
workstation, Ax = axial image, L = lung window, ST = soft-tissue window, MPR =
multiplanar reconstruction, Siem = Siemens Medical Solutions Wizard
workstation, Vital = Vital Images Vitrea workstation.
|
|
View larger version (16K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5 —Box plot shows error values (measured size minus true size)
for 5-mAs data set. Box represents 50% of values. Horizontal line = median
value, vertical line = range of values, GE = GE Healthcare Advantage
workstation, Ax = axial image, L = lung window, ST = soft-tissue window, MPR =
multiplanar reconstruction, Siem = Siemens Medical Solutions Wizard
workstation, Vital = Vital Images Vitrea workstation.
|
|
Table 1 shows the combined
accuracy of all methods for each workstation at 65 mAs according to polyp size
and shape. Table 2 shows the
improved accuracy of measurements for each workstation with optimal measuring
methods at 65 mAs for polyps of different sizes and shapes. The most accurate
methods (per mean error and SD) for each workstation at the 65-mAs dose were
2D MPR with lung window (Advantage), 3D tissue cube (Vitrea), and axial with
lung window (Wizard). For these optimal methods at 65 mAs, absolute mean error
was 0.61 mm (SD, 0.51 mm; Advantage 2D MPR with lung window), 0.76 mm (SD,
0.79 mm; Vitrea 3D tissue cube), and 0.48 mm (SD, 0.47 mm; Wizard axial with
lung window). Table 3 shows the
pairwise comparisons between these optimal measurement strategies and among
workstations. No workstation performed significantly better than another
(p = 0.11-0.60). Figure
4 shows that true polyp size was undermeasured with nearly all
methods, except for the Vitrea 3D tissue cube and the Wizard 2D MPR with lung
windows.
View this table:
[in this window]
[in a new window]
|
TABLE 2: Comparison of Workstation, Polyp Size, and Polyp Shape in Relation to
Absolute Error at 65 mAs for Best Methods
|
|
Accuracy of measurements also varied with intrinsic features of polyps
(size and shape) at the 65-mAs dose setting. Across all measurement
strategies, pedunculated polyps were more accurately measured than flat polyps
(p = 0.01); however, there was no significant difference in
measurement accuracy between pedunculated and flat polyps when optimal
measurement strategies were used (Table
2). No statistical differences were seen in pairwise comparisons
of other shapes. When all methods of measurement were considered, smaller
polyps were associated with more accurate measurements (12-mm polyps vs 10-mm
polyps, p = 0.05; 10 mm vs 7 mm, p = 0.71; 10 mm vs 5 mm,
p = 0.0005; 7 mm vs 5 mm, p = 0.001). At optimal imaging
methods, 12-mm polyps also were measured less accurately than smaller polyps
(12 mm vs 10 mm, p = 0.05; 12 mm vs 7 mm, p = 0.07; 12 mm vs
5 mm, p = 0.003; 10 mm vs 7 mm, p = 0.7; 7 mm vs 5 mm,
p = 0.17).
For the 5-mAs data set, the most accurate methods (per mean error and SD)
for each workstation at the 5-mAs dose were 2D MPR with lung windows
(Advantage), 2D MPR with lung windows (Vitrea), and axial with lung windows
(Wizard). For these optimal methods at 5 mAs, mean absolute error was 0.51 mm
(SD, 0.55 mm; Advantage 2D MPR with lung windows), 0.82 mm (SD, 0.67 mm;
Vitrea 2D MPR with lung window), and 0.57 mm (SD, 0.45 mm; Wizard axial with
lung window). Table 3 shows the
pairwise comparisons between these optimal measurement strategies and among
workstations at this very low dose. At the 5-mAs dose, the Advantage 2D MPR
with lung window and Wizard 2D axial with lung window methods were more
accurate than the Vitrea 2D MPR with lung window method (p = 0.015
and p = 0.017, respectively). Like the observations for 65 mAs, true
polyp size was undermeasured with nearly all methods, except for the Vitrea 3D
tissue cube and the Wizard 2D MPR with lung windows.
As in the higher-dose data sets, significant differences were observed in
measurement accuracy between polyps of different shapes and sizes. When all
methods of measurement were combined (Table
1), measurements of pedunculated polyps were more accurate than
those of flat polyps (p = 0.008); however, there was no significant
difference in measurement accuracy between pedunculated and flat polyps for
optimal measurement strategies. Significant differences in accuracy between
other shape combinations did not exist (p = 0.07-0.41). As was the
case for the 65-mAs data set, 5-mm polyps were measured most accurately and
with significant differences compared with the 12-mm polyps (p =
0.01, all methods; p = 0.21, optimal methods) and with the 10-mm
polyps (p = 0.02, all methods; p = 0.70, optimal
methods).
Overall the intraclass correlations ranged from 0.07 to 0.46 (slight
agreement to moderate agreement). These ranges for the Advantage, Vitrea, and
Wizard workstations were 0.23-0.33, 0.07-0.46, and 0.10-0.19, respectively.
The average absolute difference in measurements between each pair of reviewers
was 0.62 mm (0.59 SD) with optimized measurement strategies and 1.47 mm (1.14
SD) for all measurement strategies pooled.
Discussion
Because polyp size correlates with risk of malignancy and influences the
decision to recommend endoscopic polypectomy, it is imperative that
radiologists accurately measure polyp size on CTC workstations. Our study
showed that although polyps can be measured with a variety of 2D and 3D images
at various window settings, there are measurement methods for each workstation
that are significantly more accurate than others (p < 0.0001).
Furthermore, the mean error in polyp measurement was much less than 1.0 mm for
polyps of all shapes and sizes when optimal measuring methods were used on the
three workstations we evaluated. We conclude that computer workstations can be
used for accurate measurement of polyps across a range of shapes and
clinically important sizes.
Radiologists should be familiar with the optimal measurement strategies for
the CTC workstations they use. We found that significantly more accurate
measurements were obtained with 2D axial and MPR images acquired with lung
window settings than with the same or similar 2D images acquired with
soft-tissue window settings, regardless of workstation (p <
0.0001).
We found that features intrinsic to polyps themselves may affect
measurement accuracy. Pedunculated polyps were measured more accurately than
flat ones regardless of dose when all methods of measurement were combined.
This finding likely arises from the structure of these polyps, the way they
project into the colonic lumen, and the shape of pedunculated polyps in our
phantom model. The head of a pedunculated polyp projects into the colonic
lumen and is surrounded by air, so both the shape and the marked attenuation
difference at the polyp-air border help radiologists define the polyp
boundary. Conversely, the edges of a flat polyp are detected only with
morphologic clues, because both the polyp and the colonic wall are of
soft-tissue attenuation. Polyp shape did not significantly affect measurement
accuracy when optimal measurement methods were used, however, because lung
window settings presumably allowed better visualization of the edges of flat
polyps.
We found that smaller polyps were measured significantly more accurately
than larger polyps, both across methods and, considering optimal measurement
methods alone, at routine dose settings. Because optimal measurements were
obtained from 2D images on two of the workstations we studied (Advantage and
Wizard), one explanation may be that the larger polyps become, the more
difficult it may be to truly identify the greatest cross-sectional linear
measurement, because the polyp is displayed on more images. Pickhardt et al.
[12], using a Viatronix
workstation, found that linear measurements from 3D endoluminal images were
more accurate than measurements from 2D images. For the workstations we
studied, however, these sorts of inaccuracies resulting from 2D images would
be of no clinical consequence, because the absolute error resulting from these
2D measurements was less than 1.0 mm and was of the same magnitude described
by Pickhardt et al. when they used linear measurement from 3D endoluminal
views.
Our study differed in several ways from other studies examining measurement
of polyp size on CTC
[12-14].
Most important, we examined polyp measurement using three workstations.
Although automated polyp measurements may be more precise and accurate in the
future [12], our goal was to
establish optimal measurement strategies for radiologists currently practicing
CTC. Our study also included assessment of multiple reviewers and polyps of
varying sizes and shapes, so we could test for the effects of these intrinsic
polyp features on accuracy of measurement. Finally, we examined polyp
measurement using both soft-tissue and lung-window settings.
We recognize several limitations of our study. Technological assessments
are necessarily time limited, because technology continues to advance, as with
the development of colonography software for automated polyp measurement
[12,
13]. In our study the polyps
measured were uniform and symmetric. None was pancake flat (e.g., the height
of our flat polyps was one half of the width not one fourth or less, as is
observed in some cases). Elongated and asymmetric polyps with long axes
positioned obliquely to the normal axial plane would favor 3D or MPR oblique
methods for more accurate measurement. Comparing image selection (i.e., axial
slice number) for measurement accuracy may have been helpful for determining
whether this decision was the primary cause of interobserver variability.
However, this comparison could have been accomplished only for axial
measurements, and the absolute differences between reviewer measurements were
small. Furthermore, the Wizard workstation interpolates a data set into a 3D
value and does not display axial slice numbers. Our phantom model simulated
optimal conditions: No collapsed colon segments or stool interfered with our
measurements. All workstations readily allowed 2D axial, sagittal, and coronal
imaging, but the ease with which oblique planes could be created likely
resulted in more frequent use of 2D MPR measurements on some workstations than
on others. This phenomenon may have resulted in slight differences in 2D MPR
measurements among workstations but did not affect overall findings.
Although optimal methods for measuring polyp size vary among CTC
workstations, these workstations allow highly accurate measurement of polyps.
The methods do not differ significantly in facilitating accurate polyp
measurement at routine dose settings. True polyp size was undermeasured with
nearly all methods, although by a small amount. Small polyps were more
accurately measured than large polyps. Mean absolute error for the
best-performing methods for each workstation ranged from 0.48 to 0.76 mm at 65
mAs. If linear measurements of polyp size are obtained from 2D images, lung
window settings should be used.
Acknowledgments
We thank Jeff L. Fidler for his advice and suggestions about our
manuscript.
References
- Johnson CD, Harmsen WS, Wilson LA, et al. Prospective blinded
evaluation of computed tomographic colonography for screen detection of
colorectal polyps. Gastroenterology2003; 125:311
-319[CrossRef][Medline]
- Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual
colonoscopy to screen for colorectal neoplasia in asymptomatic adults.
N Engl J Med2003; 349:2191
-2200[Abstract/Free Full Text]
- Cotton PB Durvalski VL, Pineau BC, et at. Computed tomographic
colonography (virtual colonoscopy): a multicenter comparison with standard
colonoscopy for detection of colorectal neoplasia.
JAMA 2004;291:1713
-1719[Abstract/Free Full Text]
- Rockey DC, Paulson E, Niedzwiecki D, et al. Analysis of air
contrast barium enema, computed tomographic colonography, and colonoscopy:
prospective comparison. Lancet2005; 365:305
-311[Medline]
- Nusko G, Mansmann U, Partzsch U, et al. Invasive carcinoma in
colorectal adenomas: multivariate analysis of patient and adenoma
characteristics. Endoscopy1997; 29:626
-631[Medline]
- Grassi ACV, Fracasso P, Lapenta R, et al. Mediumlarge polyps of the
colon: a contribution for their clinical profile and a proper surveillance.
J Exp Clin Cancer Res1997; 16:313
-319[Medline]
- Morson B. President's address: the polyp-cancer sequence in the
large bowel. Proc R Soc Med1974; 67:451
-457[Medline]
- Church JM. Clinical significance of small colorectal polyps.
Dis Colon Rectum2004; 47:481
-485[CrossRef][Medline]
- Pickhardt P. By-patient performance characteristics of CT
colonography: importance of polyp size threshold data.
Radiology2003; 229:291
-293[Free Full Text]
- Johnson KT, Johnson CD, Anderson SM, Bruesewitz MR, McCollough CH.
CT colonography: determination of optimal CT technique using a novel colon
phantom. Abdom Imaging2004; 29:173
-176[Medline]
- Bodily KD, Fletcher JG, Engelby T, et al. Nonradiologists as second
readers for intraluminal findings at CT colonography. Acad
Radiol 2005;12:67
-73[CrossRef][Medline]
- Pickhardt PJ, Lee AD, McFarland EG, Taylor AJ. Linear polyp
measurement at CT colonography: in vitro and in vivo comparison of
two-dimensional and three-dimensional displays.
Radiology2005; 236:872
-878[Abstract/Free Full Text]
- Burling DR, McQuillan J, Honeyfield L, et al. Computed tomography
colonography: automated diameter and volume measurement of colonic polyps
compared with a manual technique: in vitro study. J Comput Assist
Tomogr 2005;29:387
-393[CrossRef][Medline]
- Blake ME, Soto JA, Hayes RA, Ferrucci JT. Automated volumetry at CT
colonography: a phantom study. Acad Radiol2005; 12:608
-613[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
E. P. Whitlock, J. S. Lin, E. Liles, T. L. Beil, and R. Fu
Screening for Colorectal Cancer: A Targeted, Updated Systematic Review for the U.S. Preventive Services Task Force
Ann Intern Med,
November 4, 2008;
149(9):
638 - 658.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. Siddiki, J. G. Fletcher, B. McFarland, N. Dajani, N. Orme, B. Koenig, M. Strobel, and S. M. Wolf
Incidental Findings in CT Colonography: Literature Review and Survey of Current Research Practice.
J. Law Med. Ethics,
June 1, 2008;
36(2):
320 - 331.
[PDF]
|
|
|
|
|
|
|
C. van Wijk, J. Florie, C. Y. Nio, E. Dekker, A. H. de Vries, H. W. Venema, L. J. van Vliet, J. Stoker, and F. M. Vos
Protrusion Method for Automated Estimation of Polyp Size on CT Colonography
Am. J. Roentgenol.,
May 1, 2008;
190(5):
1279 - 1285.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
B. Levin, D. A. Lieberman, B. McFarland, R. A. Smith, D. Brooks, K. S. Andrews, C. Dash, F. M. Giardiello, S. Glick, T. R. Levin, et al.
Screening and Surveillance for the Early Detection of Colorectal Cancer and Adenomatous Polyps, 2008: A Joint Guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology
CA Cancer J Clin,
May 1, 2008;
58(3):
130 - 160.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. G. Fletcher, F. Booya, Z. Melton, K. Johnson, L. Guendel, B. Schmidt, C. H. McCollough, B. Young, J. L. Fidler, and W. S. Harmsen
Automated Polyp Measurement with CT Colonography: Preliminary Observations in a Phantom Colon Model
Am. J. Roentgenol.,
April 1, 2007;
188(4):
945 - 952.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. J. Stanley
The Challenge of Managing Prostate Cancer
Am. J. Roentgenol.,
January 1, 2007;
188(1):
1 - 1.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
