Genitourinar y Imaging • Original Research

Metser et al.
64-MDCT of Urinary Tract Calculi

Genitourinary Imaging
Original Research
Downloaded from www.ajronline.org by 36.65.237.12 on 12/17/19 from IP address 36.65.237.12. Copyright ARRS. For personal use only; all rights reserved

Assessment of Urinary Tract

FOCUS ON:

Calculi With 64-MDCT: The

Axial Versus Coronal Plane
Ur Metser 1,2 OBJECTIVE. The objective of our study was to compare the detection rate, conspicuity,
Sangeet Ghai2 and size measurements of urinary tract calculi on coronal reformations versus the axial plane
Yang Yi Ong2 using 64-MDCT.
Gina Lockwood 3 MATERIALS AND METHODS. For this retrospective study, 80 consecutive CT ex-
Sidney B. Radomski
4 aminations performed for clinical diagnosis of renal colic or for the assessment of known
nephrolithiasis were evaluated. All studies were stripped of patient identifiers, and the axial
Metser U, Ghai S, Ong YY, Lockwood G, Radomski and coronal plane images of each study were randomized and presented to two abdominal
SB radiologists. For each study, the radiologists recorded the number and location of stones, di-
agnostic confidence and stone conspicuity (subjectively on a 2-point scale), and stone size.
The standard of reference was data from a consensus reading with the study coordinator ex-
amining the same parameters on images in both planes of each patient. Detection rates were
compared between planes using logistic regression with generalized estimating equations to
account for multiple stones per patient.
RESULTS. On consensus reading, 272 stones were identified. For all renal stones, the
coronal plane detected more stones as compared with the axial plane (p < 0.001). For stones
smaller than 5 mm, a higher proportion received the maximal conspicuity score on the coro-
nal plane than on the axial plane (p < 0.001). Both reviewers better estimated stone size on
the coronal plane than the axial plane (p = 0.02); their axial plane measurements underesti-
mated stone size by 13.4% (mean).
CONCLUSION. The detection of stones and estimation of maximal stone diameter were
Keywords: axial plane imaging, coronal plane imaging, improved using coronal reformations. The conspicuity of stones and diagnostic confidence in
MDCT, urinary tract calculi identifying stones smaller than 5 mm in diameter were also improved on the coronal plane.
DOI:10.2214/AJR.08.1545

Received July 15, 2008; accepted after revision
December 11, 2008.
1
Joint Department of Medical Imaging, Princess
Margaret Hospital, University Health Network, 610
University Ave., 3-960, Toronto, ON M5G 2M9, Canada.
Address correspondence to U. Metser
(ur.metser@uhn.on.ca).
2 enhanced CT can be used to estimate its
Joint Department of Medical Imaging, University Health
Network, Mount Sinai Hospital and Women’s College
Hospital, University of Toronto, Toronto, ON, Canada.
3
Department of Biostatistics, Princess Margaret
Hospital, Toronto, ON, Canada.
4 hanced phase may better depict urinary
Division of Urology, University Health Network,
University of Toronto, Toronto, ON, Canada.
AJR 2009; 192:1509–1513
0361–803X/09/1926–1509
© American Roentgen Ray Society
U
nenhanced CT is the imaging
technique of choice for identify-
ing renal tract calculi with sig-
nificantly improved clinical per-
formance compared with excretory urography
and a reported sensitivity and specificity of
97% and 98%, respectively, in symptomatic
patients [1]. Once a stone is identified, un-
size and location, which are important fac-
tors in determining clinical management
[2]. In addition, unenhanced CT is often in-
corporated in CT urography protocols for
assessment of hematuria because the unen-
tract calculi as compared with the contrast-
enhanced phases [3].
MDCT scanners enable rapid higher-reso-
lution imaging and improved multiplanar
reformations (MPRs) and 3D reconstruc-
tions, transforming CT from a transaxial
cross-sectional imaging technique into a
3D imaging technique [4–6]. Since 2001,
64-MDCT scanners have been in clinical
use, with near-isotropic voxels and improved
z-axis [5, 6]. The advantages of high-resolu-
tion MPRs have been described for several
clinical applications, mostly thoracic and
vascular [7–10]. In previous studies, investi-
gators have reported that coronal reforma-
tions on 64-MDCT using a 0.5-mm collima-
tion showed substantial agreement with the
axial images used to evaluate most thoracic
abnormalities [7, 8]. A previous study showed
no improvement in the detection rate of uri-
nary tract calculi when assessing thin-slice
coronal reformations obtained with a
4-MDCT scanner [11]; however, size mea-
surements were not obtained in the coronal
plane for that study. The purpose of the cur-
rent study was to evaluate the detection rate
and conspicuity of urinary tract calculi and

AJR:192, June 2009 1509


to compare stone size measurements and over-
all stone load estimates on coronal reforma-
tions versus the axial plane using 64-MDCT.
Materials and Methods
Patient Population
This retrospective study assessed 80 consecutive
Downloaded from www.ajronline.org by 36.65.237.12 on 12/17/19 from IP address 36.65.237.12. Copyright ARRS. For personal use only; all rights reserved
patients (58 men and 22 women; age range, 18–80
years; mean age, 55 years) who had been referred
to undergo CT for clinical diagnosis of renal colic
or for assessment of known nephrolithiasis. Only
one study for each patient was included.
CT Technique
All scans were obtained on an MDCT scanner
with 64 parallel detector-rows (Aquilion 64, Toshiba
Medical Systems); an individual detector width of
0.5 mm was used with a 0.5-second gantry rotation
time and a table speed of 53 mm per rotation. The
following scanning parameters were used: detector
collimation, 0.5 mm × 64; reconstruction slice
thickness, 5 mm; increment, 2.5 mm; automated
tube current modulation (mA); 120 kVp; coronal
reformation reconstruction thick­ness, 3 mm; and
increment, 3 mm. No oral or IV contrast material
was administered.
Scan Interpretation and Analysis
All studies were reviewed in the stack mode on
a PACS workstation (eFilm, Merge Healthcare).
The objective was to test the overall performance
of the axial and coronal planes for the detection of
urinary tract stones. For this purpose, all studies
were stripped of patient identifiers, and all planes of
each study were randomized and presented to two
abdominal radiologists: one abdominal imaging
fellow (reviewer 1) and one fellowship-trained staff
abdominal radiologist (reviewer 2). To minimize
recall bias, the different data sets obtained in each
patient were read at least 2 weeks apart.
For each calculus detected, radiologists
recorded the location of the calculus, number of
calculi in each location, diagnostic confidence and
stone conspicuity, and size of the stone.
Location of calculi—If a calculus was in the
kidney, the stone was localized to a calyx by
location (right or left kidney; upper, middle, or
lower pole or renal pelvis; and anterior, posterior,
medial, or lateral). If the calculus was in the ureter,
the stone was localized to the following locations:
ureteropelvic junction; upper, mid, or lower third
of ureters; or ureterovesical junction.
Number of calculi in each location—Reviewers
recorded the number of calculi in each location.
Diagnostic confidence and stone conspicuity—
Reviewers assessed diagnostic confidence and stone
conspicuity subjectively on a 2-point scale for each
stone: 1, possible calculus; or 2, definite calculus.
1510 AJR:192, June 2009
Metser et al.
TABLE 1:  Sensitivity of Each Plane for Stone Detection
Imaging Plane % of Stones Detected
All stones
Axial 84.9
Coronal 95.2
TABLE 2: Estimated Proportion of Stones with Conspicuity Score of 2
(Definite Stones) for Each Plane and Each Reader
Estimated % of Stones
Imaging Plane or Reader with Conspicuity of 2
Plane
Axial 63.7
Coronal 74.8
Reviewer
1 70.9
2 68.1
Stone size—Reviewers determined stone size
by 2D measurement. Readers were allowed to use
the zoom function to measure stones.
Standard of Reference
After completion of the independent
randomized readings, a consensus reading with
the study coordinator was performed. All data sets
of each CT study were reviewed during the same
session for the presence and location of urinary
tract stones. The largest two dimensions of a stone
on any plane on the consensus reading determined
the final stone size. A note was also made whether
the stone was vertically oriented. Data from the
consensus reading were considered the standard
of reference for independent assessment of the
axial and coronal planes.
Statistical Analysis
Statistical analysis was performed using SAS
software (version 9.1, SAS Institute); all tests
were two-sided. The sensitivity of each plane for
stone detection was calculated and tested using
logistic regression models with generalized
estimating equations (GEEs) to account for
multiple observations per patient. The effects of
plane and reviewer and of a possible interaction
were estimated using least square means. Stone
size and size accuracy were all tested using
mixed models with variance components to
adjust for multiple observations per patient. The
proportion of stones rated with a conspicuity
score of 2 was examined using logistic
regression with GEEs to allow multiple stones
per patient. Plane, reviewer, stone orientation,
and type of stone (ureteric vs renal) were tested
and estimated using least square means. Binary
interactions were also investigated.
95% CI p
< 0.001
80.8–89.0
93.1–97.3
95% CI p
< 0.001
56.1–70.7
68.4–80.2
0.10
64.5–76.6
61.0–74.5
Results
Stone Detection
On consensus reading, 272 stones were
identified: 256 renal stones and 16 ureteric
stones. Table 1 summarizes the sensitivity of
each plane for stone detection. For all stones
and for only the stones with a conspicuity score
of 2 (Table 2), the coronal plane detected more
stones than the axial plane (p < 0.001 for both;
Fig. 1). For the axial and coronal planes, the
sensitivity for the detection of stones that were
not vertically oriented in the plane was 87.5%
and 93.3%, respectively, as compared with
79% and 98%, respectively (p = 0.01 for both),
for the 84 stones found in vertically oriented
calyces. Fifteen stones were erroneously iden-
tified on either axial or coronal plane images,
as determined on consensus reading. There
was no statistically significant difference in
false-positive results between the planes.
Each kidney was also analyzed separately
for the presence or absence of calculi regard-
less of the number of calculi. There was no
significant difference in the detection of posi-
tive and negative kidneys between the planes.
However, for kidneys harboring stones with
a conspicuity score of 2, detection rates were
85.64% and 92.08% for the axial and coronal
planes, respectively (p = 0.02). Similarly, on a
patient-based analysis, the sensitivity for con-
fidently identifying patients with stones (con-
spicuity score = 2) was significantly higher for
the coronal plane than the axial plane (94.4%
and 88%, respectively, p = 0.02).
Stone Size
The average stone size as determined by
the largest diameter was significantly greater
 64-MDCT of Urinary Tract Calculi

accuracy of 0.44 vs 0.17 mm [Δ = 0.27], p <

0.001). The size of vertically oriented stones
was underestimated by 21.4% and 0.9% in
the axial and coronal planes, respectively.

Conspicuity Score
Overall, the coronal plane had significant-
Downloaded from www.ajronline.org by 36.65.237.12 on 12/17/19 from IP address 36.65.237.12. Copyright ARRS. For personal use only; all rights reserved

ly more stones classified with a conspicu-

ity score of 2 (definite stones) than the axi-
al plane (Table 2 and Fig. 2). This difference
was greater for vertically oriented stones (axi-
al, 68.1%; coronal, 85.4%; p = 0.01). Review-
ers 1 and 2 did not differ significantly in the
proportion of stones with a conspicuity score
of 2 (p = 0.10). When analyzing only stones
A B with a maximal diameter larger than 5 mm
Fig. 1—Stone detection is improved on coronal reformation in 26-year-old man who presented with left flank pain. (n = 45), no significant difference in the pro-
A, No stones were identified in left kidney by either reviewer on this axial CT scan. portion of stones with a conspicuity score of
B, On coronal reformation, both reviewers identified small stone (arrow).
2 was identified (95.6%, 98.9% for the axial
and coronal planes, respectively; p = 0.23).
for the coronal plane than the axial plane dimension was underestimated more in the
(4.27 and 3.81 mm, respectively; p < 0.001). axial plane than the coronal plane (13.4% vs Discussion
Similarly, stone area, as calculated by the 2.9%, respectively). The difference between Our study results show that the coronal
product of the two perpendicular measure- the planes in stone size measurements was and axial planes are complementary for the
ments on either plane, was significantly larg- larger for reviewer 1 (p = 0.02). Reviewer 2 detection of stones in the urinary tract be-
er for the coronal than the axial plane (86.9 more accurately estimated both stone size cause stones are missed in either plane when
and 76 mm2, respectively; p < 0.001). and area (p < 0.001 and 0.003, respectively) evaluated alone. The coronal plane was es-
The accuracy of maximal stone size and than reviewer 1. The difference in stone size pecially important in confidently identifying
of stone area on both planes was determined. measurements was also seen for stones ≥ 5 stones and in accurately assessing the size of
The mean difference of stone size and of mm: mean maximum dimension, 7.09 mm stones oriented in the vertical plane. In fact,
stone area for any plane was calculated by for the axial plane and 7.80 mm for the cor- for all stones, the average underestimation of
subtracting the maximal diameter and area of onal plane (p < 0.001). For this subgroup as stone size was more than 13% in the axial
a stone on consensus reading from the orig- well, the coronal plane underestimated size plane compared with less than 3% in the cor-
inal reviewers’ independent measurements. less than the axial plane (p < 0.001). This onal plane. These data are concordant with
Both reviewers better estimated stone size on difference in stone size accuracy was larger previously published data that showed stone
the coronal plane than on the axial plane (p < for vertically oriented stones (estimated ac- size was underestimated by 12% when CT in
0.001); they underestimated stone size on curacy of 0.94 vs 0.04 mm [Δ = 0.9]) than the axial plane was assessed compared with
the axial plane. On average, the maximum for nonvertically oriented stones (estimated an abdominal radiograph [12]. We found
that underestimation of stone size in the axi-
al plane for vertically oriented stones can ex-
ceed 20%.
It is not surprising that stone conspicui-
ty on the coronal plane is improved only for
small stones (< 5 mm) and that larger stones
are confidently depicted on either plane. Al-
though the clinical impact of confidently
identifying a small nonobstructing renal cal-
culus may be questionable, there may be a
few clinical scenarios in which the improved
stone conspicuity and stone detection rate in
the coronal plane may be advantageous.
Unenhanced CT is routinely used to eval-
uate patients with acute flank pain suspect-
ed of having renal colic. In the absence of
A B a urinary tract calculus and if no alternative
Fig. 2—Conspicuity of stone is improved on coronal reformation in 57-year-old man with history of right abdominopelvic cause for the acute symp-
ureteric stone.
A, Axial CT scan shows questionable stone (arrow) that both reviewers assigned conspicuity score of 1. toms is identified, further investigations may
B, Both reviewers confidently identified stone (conspicuity score = 2) (arrow) on coronal reformation. be necessary. Similarly, further diagnostic

AJR:192, June 2009 1511


workup or imaging surveillance may be nec-
essary for patients evaluated for hematuria if
no cause is identified. In our study population,
20% of stones were confidently identified in
only one plane. Nonobstructing renal calculi
may be clinically significant, as recently sug-
gested by investigators. In one study, Furlan
Downloaded from www.ajronline.org by 36.65.237.12 on 12/17/19 from IP address 36.65.237.12. Copyright ARRS. For personal use only; all rights reserved
et al. [13] reported that unenhanced CT iden-
tifies small nonobstructing stones on the ip-
silateral side of flank pain in up to 18% of pa-
tients seen in an emergency department for
suspected renal colic. In keeping with previ-
ously published reports of symptomatic re-
lief after removal of such small nonobstruct-
ing calculi, they concluded that although not
usually recognized as the cause of pain, in
the absence of other clinical or CT evidence
of a separate cause, these small calculi may
be responsible for a patient’s acute symptoms
and may be responsible for multiple visits to
an emergency department [13]. Investigators
believe that such small calculi may intermit-
tently obstruct the ureteropelvic junction or
the infundibulum–calyceal level, potentially
causing pain and uroepithelial damage with
hematuria [14, 15].
Little has been published about the nat-
ural history of asymptomatic renal calculi.
Glowacki et al. [16] reported their experi-
ence with 107 patients with asymptomatic re-
nal calculi. Almost a third of patients (32%)
became symptomatic within 32 months, and
more than half of them required intervention
such as extracorporeal shock wave lithotrip-
sy (ESWL), ureteroscopy, or percutaneous
nephrolithotomy [16]. Although the natural
history of small renal calculi also has not
been fully elucidated, small stone fragments
under 5 mm in diameter may be clinically
significant after ESWL. Other investigators
have suggested that although patients with
small non-infection-related stone fragments
after ESWL may be followed up expectant-
ly, a significantly number of patients (> 40%)
will require intervention or will have symp-
tomatic episodes within 2 years of ESWL
[17, 18]. Although not yet assessed in this pa-
tient population, preliminary data from our
study imply that the coronal plane could be
beneficial for improved detection and follow-
up of these potentially significant small stone
fragments. This hypothesis, however, needs
to be addressed in a separate clinical trial.
Our study results are in contradiction to
those of a previously published study [11]: In
that study, urinary tract stone detection rates
were compared for axial images and coronal
reformations obtained on a 4-MDCT unit.
1512 AJR:192, June 2009
Metser et al.
A B
Fig. 3—Stone size is better estimated on coronal reformation in 18-year-old woman with right flank pain.
A, Axial CT scan shows right distal ureteric stone (arrow) measuring 5 mm in maximal diameter.
B, On coronal reformation, vertically oriented stone (arrow) measures 11 mm in maximal diameter.
The previous study showed no added val-
ue for coronal reformations in detecting re-
nal tract calculi. The improved performance
of 64-MDCT seen in our study can be ex-
plained by several factors. 64-MDCT per-
formed with a field of view of 320 mm and
matrix size of 512 × 512 yields an in-plane
voxel dimension of approximately 0.625
mm. The source voxel dimensions for a 64-
MDCT coronal reformatted images recon-
structed from 0.5-mm axial images are near-
ly isotropic: 0.625 × 0.625 × 0.5 mm. Image
quality is further enhanced with 64-MDCT
because of reduced motion artifacts (re-
duced image acquisition time) and reduced
cone beam artifacts with the use of 3D back-
projection image reconstruction algorithms
[8]. We believe that the near-isotropic voxels
obtained with 64-MDCT result in improved
image quality of coronal reformations and
improved stone detection.
We acknowledge limitations to this retro-
spective study. First, no true standard of ref-
erence exists for the presence of renal tract
calculi. However, CT is the most sensitive
technique for detecting urinary tract calcu-
li. The combined data from both axial and
coronal planes as agreed on by a forum of
radiologists seemed like the best possible
compromise. Although no in vivo stone size
measurements could be obtained, our study
results showing underestimation of stone
size by 13.4% in the axial plane are concor-
dant with those of a previous study compar-
ing stone size as measured on axial CT and
an abdominal radiograph [12].
Second, despite subjective differences in
conspicuity scoring, for both reviewers le-
sion conspicuity in the coronal plane was fur-
ther enhanced for vertically oriented calculi.
Whether this information will positively im-
pact management of patients with renal tract
calculi remains to be determined clinically.
Third, the number of ureteric stones in
this study was relatively small. Nevertheless,
many ureteric stones are vertically oriented
and the improved estimated size in the coro-
nal plane, as suggested by our study results,
may be clinically significant (Fig. 3). Ureteric
calculi smaller than 5 mm have a more than
98% chance of spontaneous passage, whereas
calculi between 5 and 7 mm and those larger
than 7 mm in maximal diameter have a 60%
and 39% chance of spontaneous passage, re-
spectively [19, 20]. Therefore, in assessing the
chance of spontaneous passage of a stone and
in determining therapeutic strategies, accu-
rate determination of stone size is crucial.
In conclusion, for urinary tract calculi
smaller than 5 mm in diameter, there is im-
proved stone conspicuity and detection rate
in the coronal plane when assessed with
64-MDCT. For all stones, stone size estima-
tion is improved in the coronal plane, and
this difference is significantly more pro-
nounced for vertically oriented stones, which
can be underestimated in the axial plane by
more than 20%. Evaluation of the coronal
plane may be clinically valuable for patients
with suspected renal tract calculi if the axial
plane is negative for stones or if a stone is not
confidently diagnosed on the axial plane. It
also may be helpful in obtaining more accu-
rate stone size measurements, especially for
vertically oriented stones, before making a
decision about treatment strategies.

References
1. Worster A, Preyra I, Weaver B, et al. The accuracy
of noncontrast helical computed tomography ver-
sus intravenous pyelography in the diagnosis of
suspected acute urolithiasis: a meta-analysis. Ann
Emerg Med 2002; 40:280–286
Downloaded from www.ajronline.org by 36.65.237.12 on 12/17/19 from IP address 36.65.237.12. Copyright ARRS. For personal use only; all rights reserved
2. Preminger GM, Tiselius HG, Assimos DG, et al.;
EAU/AUA Nephrolithiasis Guideline Panel. 2007
guideline for the management of ureteral calculi.
J Urol 2007; 178:2418–2434
3. Van Der Molen AJ, Cowan NC, Mueller-Lisse UG,
Nolte-Ernsting CC, Takahashi S, Cohan RH; CT
Urography Working Group of the European Society
of Urogenital Radiology (ESUR). CT urography:
definition, indications and techniques—a guideline
for clinical practice. Eur Radiol 2008; 18:4–17
4. Reiser MF, Takashi M, Modic M, Bruenig R, eds.
Multislice CT. Berlin, Germany: Springer-Verlag,
2001
5. Flohr TG, Schaller S, Stierstorfer K, Bruder H,
Ohnesorge BM, Schoepf UJ. Multi-detector row
CT systems and image-reconstruction techniques.
Radiology 2005; 235:756–773
6. Prokop M. General principles of MDCT. Eur J
Radiol 2003; 45[suppl 1]:S4–S10
7. Nishino M, Kubo T, Kataoka ML, Gautam S,
Raptopoulos V, Hatabu H. Evaluation of thoracic
AJR:192, June 2009 1513
64-MDCT of Urinary Tract Calculi
abnormalities on 64-row multi-detector row CT:
comparison between axial images versus coronal
reformations. Eur J Radiol 2006; 59:33–41
8. Nishino M, Kubo T, Kataoka ML, Raptopoulos V,
Hatabu H. Coronal reformations of the chest on
64-row multi-detector row CT: evaluation of im-
age quality in comparison with 16-, 8- and 4-row
multi-detector row CT. Eur J Radiol 2006;
59:231–237
9. Wang XM, Wu LB, Sun C, et al. Clinical applica-
tion of 64-slice spiral CT in the diagnosis of the te-
tralogy of Fallot. Eur J Radiol 2007; 64:296–301
10. Albrecht T, Meyer BC. MDCT angiography of
peripheral arteries: technical considerations and
impact on patient management. Eur Radiol 2007;
17[suppl 6]:F5–F15
11. Memarsadeghi M, Schaefer-Prokop C, Prokop M,
et al. Unenhanced MDCT in patients with sus-
pected urinary stone disease: do coronal reforma-
tions improve diagnostic performance? AJR 2007;
189:329; [web]W60–W64
12. Dundee P, Bouchier-Hayes D, Haxhimolla H,
Dowling R, Costello A. Renal tract calculi: com-
parison of stone size on plain radiography and
noncontrast spiral CT scan. J Endourol 2006;
20:1005–1009
13. Furlan A, Federle MP, Yealy DM, Averch TD,
Pealer K. Nonobstructing renal stones on unen-
hanced CT: a real cause for renal colic? AJR 2008;
190:388; [web]W125–W127
14. Brannen GE, Bush WH, Lewis GP. Caliceal cal-
culi. J Urol 1986; 135:1142–1145
15. Mee Sl, Thuroff JW. Small caliceal stones: is ex-
tracorporeal shock wave lithotripsy justified? J
Urol 1988; 139:908–910
16. Glowacki LS, Beecroft ML, Cook RJ, et al. The
natural history of asymptomatic urolithiasis. J
Urol 1992; 147:319–321
17. Streem SB, Yost A, Mascha E. Clinical implica-
tions of clinically insignificant store fragments
after extracorporeal shock wave lithotripsy. J Urol
1996; 155:1186–1190
18. Tan YH, Wong M. How significant are clinically
insignificant residual fragments following litho-
tripsy? Curr Opin Urol 2005; 15:127–131
19. Segura JW, Preminger GM, Assimos DG, et al.
Ureteral Stones Clinical Guidelines Panel sum-
mary report on the management of ureteral cal-
culi. The American Urological Association. J
Urol 1997; 158:1915–1921
20. Dalrymple NC, Verga M, Anderson KR, et al. The
value of unenhanced helical computerized tomog-
raphy in the management of acute flank pain. J
Urol 1998; 159:735–740