Review

CT volumetry of the liver: Where does it stand

in clinical practice?
M.C. Lim
a,
*
, C.H. Tan
a
, J. Cai
b
, J. Zheng
b
, A.W.C. Kow
c
a
Department of Diagnostic Radiology, Tan Tock Seng Hospital, 11 Jalan Tan Tock Seng, Singapore 308433, Singapore
b
School of Computer Engineering, Nanyang Technological University, Block N4 Nanyang Avenue #02a-32,
Singapore 639798, Singapore
c
University Surgical Cluster, National University Hospital, 5 Lower Kent Ridge Road, Singapore 119074, Singapore
article i nformation
Article history:
Received 2 October 2013
Received in revised form
14 December 2013
Accepted 17 December 2013
Imaging-based volumetry has been increasingly utilised in current clinical practice to obtain
accurate measurements of the liver volume. This is particularly useful prior to major hepatic
resection and living donor liver transplantation where the size of the remnant liver and liver
graft, respectively, affects procedural success and postoperative mortality and morbidity. The
use of imaging-based volumetry, with emphasis on computed tomography, will be reviewed.
We will explore the various technical factors that contribute to accurate volumetric mea-
surements, and demonstrate how the accuracies of these techniques are inuenced by their
methodologies. The strengths and limitations of using anatomical imaging to estimate liver
volume will be discussed, in relation to laboratory and functional imaging methods of
assessment.
2014 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction
Liver volume estimation is undertaken in the preopera-
tive assessment of patients undergoing liver resection or
liver transplantation. In the assessment of suitability for
surgery, key considerations include preoperative baseline
liver function, patient size, standardized liver volume (SLV),
and postoperative residual liver volume (future liver
remnant or FLR). These factors are also applicable, in the
appropriate context, for a subset of patients who may
require portal vein embolization (PVE) to increase FLR vol-
ume. Volumetry may also be important for post-therapy
assessment, such as following liver transplant to assess
graft regeneration and treatment response assessment of
liver malignancies.
Volumetric determination should be a multidisciplinary
approach. The need for close communication between the
surgeon and the radiologist is vital in the determination of
the choice of surgical plane (such as hemi-hepatectomy
versus extended hemi-hepatectomy), assessment of
resectability and the visualization of tumour extent. Prior
careful assessment of the liver function is also vital, as a
diseased liver (e.g., steatosis and cirrhosis) requires signif-
icantly more residual volume as compared to a normal
healthy liver.
Computed tomography (CT) volumetry (CTV) has been
widely used as a method for the preoperative volumetric
assessment of the liver, for the indications as described
above. The use of other imaging methods, such as magnetic
resonance imaging (MRI) and ultrasound, have also been
explored and have shown reliable organ volume measure-
ments when the appropriate scanning protocols are
* Guarantor and correspondent: M.C. Lim, Tan Tock Seng Hospital, 11 Jalan
Tan Tock Seng, Singapore 308433, Singapore. Tel.: 65 63578111; fax: 65
63578158.
E-mail address: chitin86@yahoo.com.sg (M.C. Lim).
Contents lists available at ScienceDirect
Clinical Radiology
j ournal homepage: www. cl i ni cal radi ol ogyonl i ne. net
0009-9260/$ e see front matter 2014 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.crad.2013.12.021
Clinical Radiology 69 (2014) 887e895
employed. The strengths and limitations of the various
imaging modalities for liver volume estimation are dis-
cussed subsequently.
Clinical applications of CTV
Liver resection
It is imperative to achieve accurate determination of the
liver volume, especially in patients with chronic liver dis-
ease or cirrhosis where the size of the remnant liver be-
comes even more important as a prognostic factor. As the
presence of underlying liver disease can potentially inu-
ence the surgical resectability of a lesion, accurate deter-
mination of any hepatic parenchymal disease (particularly
cirrhosis) is therefore vital. There are many non-invasive
methods available in the evaluation of liver cirrhosis, such
as ultrasonic transient elastography (Fibroscan).
1
However,
histological diagnosis via liver biopsy remains the reference
standard for the diagnosis of liver cirrhosis, and should be
considered in the subset of patients with equivocal labo-
ratory and imaging ndings.
The FLR-to-SLV ratio is used as an indicator in predicting
the likelihood of postoperative liver failure after major he-
patic resection, particularly in patients with pre-existing
chronic liver disease. The SLV is based upon the regression
analysis of normal population (typically transplant donors),
and in which a formula can be calculated either from a
patients body weight (BW) or body surface area (BSA). A
study of 301 extended right hepatectomies demonstrated
an inverse correlation amongst small (<20%), intermediate
(20e30%), large (>30%) FLR volumes and increasing risk for
postoperative deaths.
3
In patients with normal livers, an SLV <20% following
major surgical resection has been found to be associated
with higher postoperative morbidity and liver insufciency,
including the length of stay in the intensive care unit.
4
The
exact FLR can be patient-specic and a range of cut-off
percentages have been proposed in various publications.
For example, in a prospective study, Ferrero and colleagues
found that an FLR of approximately 26.5% is required for
patients with a healthy liver.
5
However, for patients with
underlying liver disease, it is generally accepted that the FLR
required is considerably larger than those with a normal
liver given the impaired baseline function of the hepato-
cytes. In order to ensure surgical success and to reduce
signicant morbidity and mortality, a patient with cirrhosis
will require an FLR of >50% whereas the requirement is
>40% inpatients with high-grade steatosis.
2,5,6
To overcome
the potentially low FLR after liver resection, PVE can be
performed preoperatively to induce contralateral hyper-
trophy and, therefore, reduce the loss of liver mass
following surgery.
2
Postoperative infection is also a major cause of mortality.
Increased risk of severe infection is inversely correlated
with FLR.
7
In the study of Schindl et al.
7
on 104 patients who
underwent liver resection, analysis of the subgroup of pa-
tients with smaller relative residual liver volume showed a
signicant relation between severe hepatic dysfunction and
infection, suggesting that there may be a relationship be-
tween liver function, innate immunity, and susceptibility to
infections.
Living donor liver transplantation
The liver volume is also a key factor in the selection of
the appropriate individual for living donor liver trans-
plantation (LDLT). Imaging of the recipient should be car-
ried out as close to the time of planned transplantation as
possible so as to obtain an accurate reection of the re-
cipients disease state, particularly if there is underlying
malignancy as tumour can rapidly progress. Any vascular
invasion or thrombosis should be readily identied, as this
inuences the plausibility of transplantation. Additionally,
proximity of the tumour to the main hepatic and portal
vasculature as well as the central bile ducts should be
highlighted. In patients with end-stage cirrhosis, imaging
should actively seek to exclude hepatocellular carcinoma
(HCC). CT of the thorax and abdomen should be acquired at
the same sitting to rule out metastatic disease or concom-
itant extra-hepatic primary malignancy, should a hepatic
tumour be present.
Although the imaging ndings of the recipient have no
signicant impact on donor selection, precise assessment
of the donor liver volume is crucial in determining
whether the donor is suitable for LDLT to ensure safety for
both donor and recipient. Preoperative imaging is required
to ensure there is no underlying focal or diffuse liver dis-
ease that may make transplantation unsuitable, such as
steatosis, cirrhosis, and focal benign or malignant
neoplasms.
For accurate liver volume estimation, a good under-
standing of intrahepatic vascular and biliary anatomy is
important. Sound knowledge of the surgical procedure is
required for accurate evaluation of the donor liver volume.
The key anatomical variants that may potentially inuence
the surgical techniques should be highlighted.
8
For
example, typical anatomy of the hepatic arterial is only seen
in 55e61% of the population. Common variants include
replaced left hepatic artery from the left gastric artery and
replaced right hepatic artery from the superior mesenteric
artery as well as accessory right or left hepatic arteries.
9
Precise details of the vascular anatomy and its associated
territories can be obtained via angiography or through the
use of personalized computer analysis software, such as the
LiverAnalyzer (MeVis Distant Services, MeVis Medical So-
lutions, Bremen, Germany; Fig 1).
Using CTV there is generally good correlation of the
estimated volume with graft weight obtained.
10
A study by
Nakayama et al.
11
showed that the mean weight of an adult
liver was 881.1 249.8 g, whereas the mean measured
volume of the liver was 956.99 280.1 cm
3
.
11
For an adult donor, a remnant liver volume of 30% for the
donor is considered to be the minimum threshold for
transplantation to proceed, providing that there is no
steatosis or other underlying liver disease.
12
Small-for-size
syndrome occurs when the graft size is too small for the
M.C. Lim et al. / Clinical Radiology 69 (2014) 887e895 888
recipient, and may manifest in the form of postoperative
cholestasis, poor bile production, refractory ascites, and
prolonged prothrombin time. A minimum graft-to-
recipient weight ratio of 0.8 has been recommended by
some authors to prevent small-for-size syndrome, and the
use of volumetry calculator will help to determine the ad-
equacy of the potential liver graft.
13
However, it is important to note that graft function and
survival are not only inuenced by graft size, but can also be
affected by pre-existing disease in the donors liver prior to
transplantation. Specically, hepatic steatosis is a common
nding in developed countries and can have signicant
impact on surgical outcome in the transplant. In the study of
DAlessandro et al.
14
involving 124 donor livers for trans-
plantation, there was a signicantly higher rate of primary
non-function and initial poor function in patients who were
shown to have received livers with severe hepatic steatosis
on donor liver biopsy, compared to those who received
livers with none or lesser degrees of hepatic steatosis.
14
Upon identication of a potentially steatotic liver in the
donor, MRI of the liver can help to further quantify the
degree of steatosis.
Pre-operative increase of remnant liver volume
Transarterial chemoembolization (TACE) involves the
administration of a chemotherapeutic agent into the
hepatic artery followed by hepatic artery embolization
(HAE). TACE has been shown to be effective in downsizing
HCCs as these tumours preferentially obtain their blood
supply from the hepatic artery. Occlusion of the hepatic
artery results in selective ischaemia of the tumour and en-
hances the cytotoxic effect of the chemotherapeutic agent.
TACE is frequently performed before PVE in patients with
HCCs to prevent tumour growth during the time interval of
PVE and scheduled resection (Fig 2).
15
In addition, TACE
blocks the arterio-portal shunts that are commonly
encountered in HCC whose presence could potentially
attenuate the effects of PVE.
PVE causes reduction in the size of the embolized he-
patic lobe and hence induces compensatory hypertrophy
of the remnant portion of the liver (Fig 3). PVE has been
widely regarded as an effective means to increase the FLR
volume, particularly in patients with inadequate remnant
liver volume who require major or extended hepatec-
tomy.
16,17
This implies that a higher proportion of patients
with previously unresectable disease can now benet
from surgery following a successful PVE procedure.
18
Other than just increasing the FLR volume, PVE per-
formed prior to resection has been shown to decrease the
peri-operative mortality.
19,20
The increase in FLR after PVE
may also be used for prognostication. An increase of >5%
in FLR may indicate low risk for liver failure after major
resection.
21
Figure 1 (a) Depiction of hepatic vascular anatomy (using LiverAnalyzer). Image on the left shows hepatic arteries and portal veins whereas
image on the right shows the hepatic veins. (b) Proposed resection of the right lobe graft, which includes the territory supplied by the right
middle hepatic vein (using LiverAnalyzer). The estimated graft weight is 717 g.
M.C. Lim et al. / Clinical Radiology 69 (2014) 887e895 889
Emerging techniques
The use of HAE after PVE has also been recently reported
as a method to accelerate the hypertrophy of the remnant
liver in patients with an inadequate increase of the remnant
liver volume following unsuccessful PVE.
22
The effects of
such arterial embolization on liver regeneration were,
however, inconsistent and this procedure runs the risk of
severe liver ischaemia predisposing to hepatic abscess
formation.
23
In view of the variable results and risk of hepatic infec-
tion following HAE,
23
hepatic vein embolization (HVE) has
also been performed in patients with inadequate remnant
liver volume after PVE. In a paper by Hwang et al.,
24
the use
of HVE following PVE has shown favourable results.
24
However, limited literature is available on this emerging
technique, and more studies will be required to accurately
assess the efcacy of this procedure.
Factors affecting accuracy of CTV: technical
considerations
Phase of contrast enhancement: does it matter?
Volumetric assessment is dependent on the scan phase.
There are signicant differences in the total liver volume
Figure 2 Selected axial CT images of the liver before (a,b) and 10 weeks following (c,d) TACE of the tumour in the right lobe and PVE. There is
hypertrophy of the left hepatic lobe. Streak artefacts are noted in the proximal right portal vein (d) as a result of embolization coils.
Figure 3 PVE. Angiography image (left) demonstrates the portal vein anatomy. The catheter is seen within a segmental branch of the right portal
vein. Post-embolization image (right) shows coils within the proximal right segmental portal vein with occlusion of contrast medium ow
distally.
M.C. Lim et al. / Clinical Radiology 69 (2014) 887e895 890
and graft-liver volume calculations between the largest
(venous) and smallest (without contrast medium adminis-
tration, also known as plain) CT phases. CTV based on the
largest (venous) CT phase is associated with considerable
overestimation, whereas volumetry based on the smallest
(plain) CT phase yields the least signicant overestimation.
However, even though CTV based on the smallest (plain)
CT phase most accurately matches the actual intraoperative
ndings, the venous phase is typically preferred in prac-
tice.
25
The reason for this is because contrast-enhanced CT
delineates the vascular anatomy and the margins of any
tumour better than unenhanced CT of the liver.
Manual versus semi-automated versus automated
methods of CTV: which is better?
CTV has traditionally been performed by manual contour
tracing of the hepatic contours and summation of the liver
area on each axial section, usually by a radiologist (Fig 4).
Contour tracing is typically carried out using the standard
optical mouse. Novel ways of tracing the liver margins have
been devised, including the use of a freehand electromag-
netic pen tablet.
26
There is comparable accuracy and pre-
cision in these two methods. The mean segmentation time
per patient is signicantly shorter with the use of the
freehand electromagnetic pen contour-tracing method.
However, such manual methods are operator-dependent
and require considerable time and attention. To further
speed up the process and avoid tedious operations, auto-
mated and semi-automated ways of volumetric measure-
ments have been proposed.
11,27e29
Various techniques and algorithms have been described
for the automated methods of volumetry, including three-
dimensional (3D) active contour segmentation
27
and
geodesic active contour segmentation coupled with level
set algorithms.
11
One such technique involves the estima-
tion of the mean CT attenuation value of the liver and se-
lection of the initial candidate regions of the liver on each
section using the estimated CT value of the liver. The liver is
then separated from other adjacent organs via the analysis
of edge information inside the initially estimated liver. A 3D
image of the liver is subsequently reproduced for the
calculation of liver volume.
29
Automatic segmentation often fails for certain CT images
that are lowin contrast or have missing edges due to similar
intensity of adjacent organs or machine noise. Thus, semi-
automated methods are proposed to include the user in
the processing loop to guide automated segmentation,
which provides more exibility and control over the volu-
metric determination. An example using the random-
walk technique is shown in Fig 5.
30
Semi-automatic seg-
mentation introduced by Wimmer et al.
31
received the
highest average performance score of 84.6 as stated in the
2007 proceedings of the Medical Image Computing and
Computer-Assisted Intervention workshop regarding seg-
mentation of the liver. Basically, their algorithm can be
divided into three steps. The rst step is manual delinea-
tion. Manual delineation works on two-dimensional (2D)
multiplanar reconstruction views. It captures the cross-
section of the liver by user input control points or live
wire.
32
Experiments show that typically six to eight con-
tours can achieve good results. The second step is 3D sur-
face reconstruction, which works by using radial basis
function based on the manual delineation.
33
The last step is
surface evolution, which evolves the reconstructed 3D
surface using a level set framework comprising both image
and shape information.
34
In general, semi-automatic
methods have been show to outperform automatic
methods.
In terms of duration of examination, the average user
time for automated volumetry is around 0.57 0.06 min/
case, whereas those for semi-automated and manual volu-
metry are around 27.3 4.6 and 39.4 5.5 min/case,
respectively.
26
This implies a highly effective time-saving
measure of >30 min/case if one switched from manual to
automated means of volumetric assessment.
Image section thickness
As accurate liver volume estimation relies on precise
segmentation of the liver, one can expect more accurate
results when using CT images that have smaller section
thicknesses. Furthermore, there is an inverse relationship
between the section thickness and the calculated liver
volumes, that is, liver volumes obtained from analysis of
submillimetre thick CT sections are typically larger than
volume calculated from thicker images.
35
A maximumerror
of 5% in the calculated graft volume should be expected if
6 mm thick sections are used.
However, the use of thinner sections translates to a
greater number of axial sections needed for segmentation,
hence more time required for volumetric analysis. This is
Figure 4 Selected axial sections through the liver demonstrate manual contour tracing of the left hepatic lobe using syngo MultiModality
Workplace (MMW). The MMW then generates a volumetric measurement of the area of interest.
M.C. Lim et al. / Clinical Radiology 69 (2014) 887e895 891
most time consuming when a manual method of contour
tracing is utilized. Supposing that a typical adult liver span
is 12 cm, the use of 6 mm and 0.625 mm sections requires
segmentation of approximately 20 images and 192 images,
respectively. The balance between accuracy of volume
estimation and time required for image analysis should be
carefully weighed.
With regard to the use of volumetry in MRI, the source
data are derived from dynamic 3D imaging, such as volu-
metric interpolated breath-hold examination (VIBE) and
liver acquisition with volume acquisition (LAVA). Reiner
et al.
36
suggested the use of 6 mm and 8 mm section
thicknesses for CT and MRI, respectively, for a good balance
between accuracy of calculated volume and time efciency.
Software for calculation of liver volumes
Volumetric measurements of the liver have been
described beyond the use of dedicated professional soft-
ware. Various authors have used more accessible, generic
image processing software such as OsiriX

(Apple Mac
OS),
37
Photoshop

,
38
ImageJ

,
39
to measure hepatic vol-
umes, and these have been found to correlate well with
conventional CTV (performed in institutions) as well as
with intra-operative graft volume. The use of these software
packages on a personal computer has also been explored. A
study showed small interobserver variability (up to 5%) in
the determination of total liver volumes, resection volumes,
and tumour volumes using OsiriX software by surgical
trainees when compared to CT scanner-linked Aquarius
(iNtuition

, by Terra Recon) software by a radiologist.

37
The
average time to complete volumetric determination with
OsiriX

is 19 3 min using a 5 mm section thickness.

38
Which formula to use to calculate liver volumes?
Formulas to estimate the SLV have been established us-
ing logistical regression statistical techniques. A patient
with a non-diseased liver requires a FLR-to-BWratio of 0.4
or FLR-to-TLV ratio of 20% in order to ensure safe and
appropriate hepatic resection.
40
A number of papers have
been published based on different ethnic groups.
41e47
This
subject is of particular interest among countries in East Asia
where hepatitis-induced liver cirrhosis and HCCs are prev-
alent. In such studies, preoperative measurement of the
liver volume is determined using CTV, and using the
explanted liver volume as a reference standard, a formula
(with coefcient factor) that is derived from the patients
BW (in kilograms) or BSA (in square metres) is obtained
through logistical regression analysis. To achieve accuracy,
larger numbers of patients need to be sampled, but as liver
transplantations are not common, such logistical regression
based studies are often limited by sample size.
As a result of the variations, there are signicant inter-
racial differences, and also, considerable differences
within the same ethnic group, depending on the study that
is being quoted. For example, in an adult patient of average
weight (60 kg), the estimated SLVs can range from
1024e1302 cm
3
. Assuming a FLR of 300 cm
3
, the FLR-to-SLV
ratio ranges from 23e29%. This can mean a difference be-
tween surgery and conservative management if a cut-off
value of 25% were to be used for decision-making. A few
of the more widely cited formulas are listed in Table 1.
41e47
The BW of a patient is commonly used as an index in
many of the formulas. However, this would be a challenge in
patients with ascites, where BW may uctuate signicantly
on a daily basis. Whereas other formulas use BSA as an in-
dex, which requires additional measurement of the pa-
tients height, which can be technically difcult in frail or ill
patients. Moreover, there are also different formulas to
calculate BSA, making calculation even more
cumbersome.
48,49
There is currently no consensus as to which formula is
better, that is, whether a BW-derived or BSA-derived for-
mula is more accurate in predicting the SLV. Chun et al.
40
compared patients BW and BSA as an indicator in the
Figure 5 Semi-automated method of assessing liver volumetry using the randomwalk technique. The liver is manually traced, followed by 3D
surface reconstruction and surface evolution.
Table 1
Examples of formulas for calculation of standardized liver volume.
Reference Formula
DeLand, 1968.
36
SLV 1020 BSA 220
Heinemann, 1999.
37
SLV BSA 1072.8 345.7
Vauthey, 2002.
38
SLV BW 18.51 191.80
SLV (BSA)
2
1267.28 794.41
Urata, 2002.
39
SLV (706.2 BSA 2.4)
Yoshizumi, 2003.
40
SLV (BSA)
2
772
Hashimoto, 2006.
41
SLV BSA 961.3 404.8
Fu-Gui, 2009.
42
SLV BW 11.508 334.024
Note: BSA (DuBois formula)
43
0.007184 (weight in kg)
0.425
(height in
cm)
0.725
; BSA (Mosteller formula)
44
(weight in kg height in cm)
0.5
/60.
M.C. Lim et al. / Clinical Radiology 69 (2014) 887e895 892
prediction of the SLV, and found strong correlation in both
of the indices. Additionally, BW formulas and BSA formulas
seem to demonstrate comparable ability in the assessment
of potential postoperative hepatic dysfunction.
More recently, SLV has been used in the estimation of the
size of the right or left hemi-liver graft for transplantation.
First described by Lee et al.,
50
this formula uses the di-
ameters (in millimetres) of the right main portal vein (R)
and left main portal vein (L) measured via sonography,
which can be done conveniently at the bedside. Using
Uratas SLV formula, the right and left hemi-hepatic graft
size can be obtained as follows: SLV [R
2
/(R
2
L
2
)] and
SLV SLV [L
2
/(R
2
L
2
)], respectively.
50
In paediatric patients, the preoperative estimation of the
graft volume by CTV is typically larger than the intra-
operative measured graft weight. Therefore, for younger
patients, an age-adjusted formula is required to adjust for
this error. For example, in the paper published by Kayashima
et al.,
51
the age-adjusted graft volume 70.767 (graft
volume estimated with CTV) (1.298 donor age).
Comparison of CTV to other imaging
methods
Attempts to use other imaging methods, such as ultra-
sound and MRI, for liver volumetry have also been made.
There appears to be a strong positive correlation in the
virtual liver volume measurement using CT and MRI.
52,53
MRI has the additional advantage of superior soft-tissue
contrast. This may theoretically allow for more accurate
segmentation of the liver margins, as well as delineation of
any focal hepatic tumours. A good knowledge of the
appropriate sequences is important for optimal assessment.
However, MRI is more expensive and is unsuitable for pa-
tients who are claustrophobic. Moreover, the quality of MRI
images is easily degraded by motion artefacts, which pre-
clude precise assessment.
Good concordance has been found between 3D ultra-
sound and 3D CTV; however, the use of conventional 2D
ultrasound has shown to be less precise compared to 3D
ultrasound and 3D CTV. Evaluation using ultrasound can be
notoriously inconsistent as it is heavily operator-dependent
and can often be technically challenging, such as the pres-
ence of a high-riding liver that is obscured by the ribcage.
54
At present, CTV is still the preferred imaging technique for
liver volumetry.
Other methods of liver function assessment
Clinical and laboratory assessment
The most common assessment of hepatic functional
reserve is based on a detailed clinical examination with
laboratory results (including liver function tests, coagula-
tion prole, and platelet count). Several scoring systems
could be formulated based on clinical and laboratory nd-
ings. The ChildePugh classication is frequently used to
assess the severity of cirrhosis as well as to extrapolate
mortality. It is also used to guide management and to
determinate the suitability of resection or transplant. It is
derived from ve indices (serum albumin level, serum
bilirubin level, extent of coagulopathy, presence of ascites,
and encephalopathy). Similarly, the Model for End-Stage
Liver Disease is also used in the assessment of patients for
liver transplantation. The score is derived from a linear
regression model based on the international normalized
ratio as well as serum bilirubin and creatinine levels.
Functional assessment
For patients with normal liver parenchyma, the functional
liver volume is proportional to the total liver volume
measured by CTV. However, in people with compromised
liver function, there would be discordance between func-
tional andmorphological imaging modalities. This is because
CTV, as withother methods of anatomical imaging, is focused
onmorphology, anddoes not reect liver functionaccurately.
Hence, at present, CTV is more benecial in patients with
non-cirrhotic livers, such as those with metastatic disease
requiring extended hemi-hepatectomy or transplantation.
The assessment of liver function using serum indoc-
yanogreen (ICG) level as a marker of biliary excretion is a
common method of evaluating liver function.
55
However,
the ICG excretion test gives a global assessment of the liver
function and does not reect differential function within
the liver segments. In addition, liver function is often not
distributed homogeneously, with the segments requiring
resection showing poorer function. Furthermore, the results
may be affected not just by hepatic function, but also by
hepatic blood ow.
56
Although some studies have com-
bined CTV with ICG excretion, among other parameters, it
has not been shown whether CTV alone may sufce in
predicting postoperative hepatic dysfunction.
CTV has been found to be inferior to the use of 99m-
technetium-galactosyl human serum album scintig-
raphy
57,58
and 99m-technetium-mebrofenin (Tc99m-
mebrofenin) hepatobiliary scintigraphy,
59e61
in the deter-
mination of the risk of liver failure and its related mortality
after hepatectomy through assessment of the functional
status of the remnant liver. In a multivariate analysis un-
dertaken by Dinant et al.,
61
uptake of Tc99m-mebrofenin
was the only signicant factor associated with liver fail-
ure.
61
The FLR was not found to be signicantly associated
with any of the outcome parameters.
The functional capacity of the liver may also be assessed
with 99mTc-mebrofenin single-proton-emission CT, which
provides a better idea of the differential hepatocyte func-
tion within each of the liver segments, compared to routine
scintigraphy. Therefore, there is a need for utilization of
combined functional and morphological assessments in the
evaluation of FLR for more accurate prediction of post-
operative hepatic dysfunction.
Conclusion
Imaging-based volumetry of the liver is vital in the pre-
operative planning for major hepatic resection or liver
M.C. Lim et al. / Clinical Radiology 69 (2014) 887e895 893
transplantation as well as in the determination of future
remnant liver volume. CTV is currently the preferred means
of imaging-based volumetry. As the techniques evolve and
become more user-friendly, clinicians managing patients
who require preoperative assessment of liver function
should be aware of its strengths and limitations, in order to
utilize CTV appropriately. In future, morphological volu-
metric assessment of the liver should be combined with the
use of functional imaging in order to more accurately reect
functional liver volume, to better predict postoperative liver
dysfunction in patients with impaired liver function.
Acknowledgement
The authors thank Dr Albert Low Su Chong, Singapore
General Hospital, for the MeVis images.
References
1. Nguyen-Khac E, Capron D. Noninvasive diagnosis of liver brosis by
ultrasonic transient elastography (Fibroscan). Eur J Gastroenterol Hepatol
2006;18:1321e5.
2. van den Esschert JW, de Graaf W, van Lienden KP, et al. Volumetric and
functional recovery of the remnant liver after major liver resection with
prior portal vein embolization: recovery after PVE and liver resection. J
Gastrointest Surg 2009 Aug;13:1464e9.
3. Kishi Y, Abdalla EK, Chun YS, et al. Three hundred and one consecutive
extended right hepatectomies: evaluation of outcome based on sys-
tematic liver volumetry. Ann Surg 2009;250:540e8.
4. Abdalla EK, Barnett CC, Doherty D, et al. Extended hepatectomy in pa-
tients with hepatobiliary malignancies with and without preoperative
portal vein embolization. Arch Surg 2002;137:675e80.
5. Ferrero A, Vigan o L, Polastri R, et al. Postoperative liver dysfunction and
future remnant liver: where is the limit? Results of a prospective study.
World J Surg 2007;31:1643e51.
6. Gulielmi A, Ruzzenente A, Conci S, et al. How much remnant is enough
in liver resection? Dig Surg 2012;29:6e17.
7. Schindl MJ, Redhead DN, Fearson KC, et al. The value of residual liver
volume as a predictor of hepatic dysfunction and infection after major
liver resection. Gut 2005;54:289e96.
8. Singh AK, Cronin CG, Verma AH, et al. Imaging of preoperative liver
transplantation in adults: what radiologists should know. RadioGraphics
2011;31:1017e30.
9. Covey AM, Brody LA, Maluccio MA, et al. Variant hepatic arterial anat-
omy revisited: digital subtraction angiography performed in 600 pa-
tients. Radiology 2002;224:542e7.
10. Wei L, Zhi ZJ, L u Y, et al. Application of computer-assisted three-
dimensional quantitative assessment and a surgical planning tool for
living donor liver transplantation. Chin Med J 2013;126.
11. Nakayama Y, Li Q, Katsuragawa S, et al. Automated hepatic volumetry for
living related liver transplantation at multisection CT. Radiology
2006;240:743e8.
12. Tsang LL, Chen CL, Huang TL, et al. Preoperative imaging evaluation of
potential living liver donors: reasons for exclusion from donation in
adult living donor liver transplantation. Transplant Proc
2008;40:2460e2.
13. Ben-Hain M, Emre S, Fishbein TM, et al. Critical graft size in adult-to-
adult living donor liver transplantation: impact of the recipients dis-
ease. Liver Transpl 2001;7:948e53.
14. DAlessandro AM, Kalayoglu M, Sollinger HW, et al. The predictive value
of donor liver biopsies on the development of primary nonfunction after
orthotopic liver transplantation. Transplant Proc 1991;23:1536e7.
15. Kokudo N, Makuuchi M. Current role of portal vein embolization/he-
patic artery chemoembolization. Surg Clin North Am 2004;84:643e57.
16. Poon Ronnie T, Sheung TF. Assessment of hepatic reserve for indication
of hepatic resection: how I do it. J Hepatobiliary Pancreat Surg
2005;12:31e7.
17. Tanaka H, Hirohashi K, Kubo S, et al. Preoperative portal vein emboli-
zation improves prognosis after right hepatectomy for hepatocellular
carcinoma in patients with impaired hepatic function. Br J Surg
2000;87:879e82.
18. Azoulay D, Castaing D, Krissat J, et al. Percutaneous portal vein embo-
lization increases the feasibility and safety of major liver resection for
hepatocellular carcinoma in injured liver. Ann Surg 2000;232:665e72.
19. Kubota K, Makuuchi M, Kusaka K, et al. Measurement of liver volume
and hepatic functional reserve as a guide to decision-making in resec-
tional surgery for hepatic tumors. Hepatology 1997;26:1176e81.
20. Palavecino M, Chun YS, Madoff DC, et al. Major hepatic resection for
hepatocellular carcinoma with or without portal vein embolization:
perioperative outcome and survival. Surgery 2009;145:399e405.
21. Ribero D, Abdalla EK, Madoff DC, et al. Portal vein embolization before
major hepatectomy and its effects on regeneration, resectability and
outcome. Br J Surg 2007;94:1386e94.
22. Gruttadauria S, Luca A, Mandala L, et al. Sequential preoperative ipsi-
lateral portal and arterial embolization in patients with colorectal liver
metastases. World J Surg 2006;30:576e8.
23. Gruttadauria S, Gridelli B. Sequential preoperative ipsilateral portal and
arterial embolization in patients with liver tumors: is it really the best
approach? World J Surg 2007;31:2427e8.
24. Hwang S, Lee SG, Ko GY, et al. Sequential pre-operative ipsilateral he-
patic vein embolization after portal vein embolization to induce further
live regenerative in patients with hepatobiliary malignancy. Ann Surg
2009;249:608e16.
25. Radtke A, Sotiropoulos GC, Nadalin S, et al. Preoperative volume pre-
diction in adult living donor liver transplantation: how much can we
rely on it? Am J Transplant 2007;7:672e9.
26. Perandini S, Faccioli N, Inama M, et al. Freehand liver volumetry by
using an electromagnetic pen tablet: accuracy, precision, and rapidity. J
Digit Imaging 2011;24:360e5.
27. Suzuki K, Epstein ML, Kohlbrenner R, et al. Quantitative radiology:
automated CT liver volumetry compared with interactive volumetry and
manual volumetry. AJR Am J Roentgenol 2011;197:W706e12.
28. Suzuki K, Kohlbrenner R, Epstein ML, et al. Computer-aided measure-
ment of liver volumes in CT by means of geodesic active contour seg-
mentation coupled with level-set algorithms. Med Phys
2010;37:2159e66.
29. Frericks BB, Kiene T, Stamm G, et al. CT-based liver volumetry in a
porcine model: impact on clinical volumetry prior to living donated
liver transplantation. Rofo 2004;176:252e7.
30. Yang W, Cai J, Zheng J, et al. User-friendly interactive image segmen-
tation through unied combinatorial user inputs. IEEE Trans Image
Process 2010;19:2470e9.
31. Wimmer A, Soza G, Hornegger J. Two-stage semi-automatic organ seg-
mentation framework using radial basis functions and level sets. In:
HeimannT, Styner M, vanGinnekenB, editors. 3Dsegmentationinthe clinic:
a grand challenge. Springer Berlin, Heidelberg: LNCS; 2007. pp. 179e88.
32. Falc~ao AX, Udupa JK, Samarasekera S, et al. User-steered image seg-
mentation paradigms: live wire and live lane. J Graph Models Imaging
Proc 1998;60:233e60.
33. Carr JC, Beatson RK, Cherrie JB, et al. Reconstruction and representation of
3D objects with radial basis functions. SIGGRAPH; 2001. pp. 67e76.
34. Sethian JA. Level set methods and fast marching methods. Cambridge:
Cambridge University Press; 1999.
35. Hori M, Suzuki K, Epstein ML, et al. Computed tomography liver volu-
metry using 3-dimensional image data in living donor liver trans-
plantation: effects of the slice thickness on the volume calculation. Liver
Transpl 2011;17:1427e36.
36. Reiner CS, Karlo C, Petrowsky H, et al. Preoperative liver volumetry: how
does the slice thickness inuence the multidetector computed tomog-
raphy- and magnetic resonance-liver volume measurements? J Comput
Assist Tomogr 2009;33:390e7.
37. van der Vorst JR, van Dam RM, van Stiphout RS, et al. Virtual liver
resection and volumetric analysis of the future liver remnant using open
source image processing software. World J Surg 2010;34:2426e33.
38. Lu Y, Wu Z, Liu C, et al. Hepatic volumetry with PhotoShop in personal
computer. Hepatobiliary Pancreat Dis Int 2004;3:82e5.
39. Dello SA, van Dam RM, Slangen JJ. Liver volumetry plug and play: do it
yourself with ImageJ. World J Surg 2007;31:2215e21.
M.C. Lim et al. / Clinical Radiology 69 (2014) 887e895 894
40. Chun YS, Ribero D, Abdalla EK, et al. Comparison of two methods of
future liver remnant volume measurement. J Gastrointest Surg
2008;12:123e8.
41. DeLand FH, North WA. Relationship between liver size and body size.
Radiology 1968;91:1195e8.
42. Heinemann A, Wischhusen F, Puschel K, et al. Standard liver volume in
the Caucasian population. Liver Transpl Surg 1999;5:366e8.
43. Vauthey JN, Abdalla EK, Doherty DA, et al. Body surface area and body
weight predict total liver volume in Western adults. Liver Transpl
2002;8:233e40.
44. Urata K, Kawasaki S, Matsunami H, et al. Calculation of child and adult
standard liver volume for liver transplantation. Hepatology
1995;21:1317e21.
45. Yoshizumi T, Gondolesi GE, Bodian CA, et al. A simple new formula to
assess liver weight. Transplant Proc 2003;35:1415e20.
46. Hashimoto T, Sugawara Y, Tamura S, et al. Estimation of standard liver
volume in Japanese living liver donors. J Gastroenterol Hepatol
2006;21:1710e3.
47. Fu-Gui L, Lu-Nan Y, Bo L, et al. Estimation of standard liver volume in
Chinese adult living donors. Transplant Proc 2009;41:4052e6.
48. DuBois D, DuBois EF. A formula to estimate the approximate surface
area if height and weight be known. Arch Intern Med 1916;17:863e71.
49. Mosteller RD. Simplied calcication of body-surface area. N Engl J Med
1987;317:1098.
50. Lee WC, Lee CS, Soong RS, et al. Split liver transplantation in adults:
preoperative estimation of the weight of right and left hemiliver grafts.
Liver Transpl 2011;17:93e4.
51. Kayashima H, Taketomi A, Yonemura Y, et al. Accuracy of an age-
adjusted formula in assessing the graft volume in living donor liver
transplantation. Liver Transpl 2008;14:1366e71.
52. Karlo C, Reiner CS, Stolzmann P, et al. CT- and MRI-based volumetry of
resected liver specimen: comparison to intraoperative volume and
weight measurements and calculation of conversion factors. Eur J Radiol
2010;75:e107e11.
53. Muggli D, M uller MA, Karlo C, et al. A simple method to approximate
liver size on cross-sectional images using living liver models. Clin Radiol
2009;64:682e9.
54. Lang H, Wolf GK, Prokop M, et al. 3-dimensional sonography for volume
determination of liver tumorsdreport of initial experiences. Chirurg
1997;68:675e80.
55. Chijiiwa K, Watanabe M, Nakano K, et al. Biliary indocyanine green
excretion as a predictor of hepatic adenosine triphosphate levels in
patients with obstructive jaundice. Am J Surg 2000;179:161e6.
56. Fazakas J, M andli T, Ther G, et al. Evaluation of liver function for hepatic
resection. Transplant Proc 2006;38:798e800.
57. Nanashima A, Yamaguchi H, Shibasaki S, et al. Relationship between CT
volumetry and functional liver volume using technetium-99m galac-
tosyl serum albumin scintigraphy in patients undergoing preoperative
portal vein embolization before major hepatectomy: a preliminary
study. Dig Dis Sci 2006;51:1190e5.
58. Wakamatsu H, Nagamachi S, Kiyohara S, et al. Predictive value of Tc-
99m galactosyl human serum albumin liver SPECT on the assessment
of functional recovery after partial hepatectomy: a comparison with CT
volumetry. Ann Nucl Med 2010;24:729e34.
59. de Graaf W, van Lienden KP, van Gulik TM, et al. (99m)Tc-mebrofenin
hepatobiliary scintigraphy with SPECT for the assessment of hepatic
function and liver functional volume before partial hepatectomy. J Nucl
Med 2010;51:229e36.
60. de Graaf W, van Lienden KP, Dinant S, et al. Assessment of future
remnant liver function using hepatobiliary scintigraphy in patients
undergoing major liver resection. J Gastrointest Surg 2010;14:369e78.
61. Dinant S, de Graaf W, Verwer BJ, et al. Risk assessment of post-
hepatectomy liver failure using hepatobiliary scintigraphy and CT
volumetry. J Nucl Med 2007;48:685e92.
M.C. Lim et al. / Clinical Radiology 69 (2014) 887e895 895