The development of a practical and uncomplicated predictive
equation to determine liver volume from simple linear ultrasound
measurements of the liver
Jessie T. Childs
a, *
, Kerry A. Thoirs
a, c
, Adrian J. Esterman
b, d
a
International Centre of Allied Health Research, University of South Australia, Australia
b
FACE, University of South Australia and Centre for Research Excellence in Chronic Disease Prevention, James Cook University, Australia
article info
Article history:
Received 12 October 2015
Received in revised form
16 December 2015
Accepted 27 December 2015
Available online 19 January 2016
Keywords:
Ultrasound
Liver
Volume
Equation
Measurement
abstract
This study sought to develop a practical and uncomplicated predictive equation that could accurately
calculate liver volumes, using multiple simple linear ultrasound measurements combined with mea-
surements of body size. Penalized (lasso) regression was used to develop a new model and compare it to
the ultrasonic linear measurements currently used clinically. A BlandeAltman analysis showed that the
large limits of agreement of the new model render it too inaccurate to be of clinical use for estimating
liver volume per se, but it holds value in tracking disease progress or response to treatment over time in
individuals, and is certainly substantially better as an indicator of overall liver size than the ultrasonic
linear measurements currently being used clinically.
© 2016 The College of Radiographers. Published by Elsevier Ltd. All rights reserved.
Ultrasound evaluation of the size of the liver can be useful for
clinicians to aid in the diagnosis of liver disease or to track disease
progress and response to treatment/intervention over time.
1,2
Ul-
trasound is a readily available, inexpensive and comparably fast
imaging technology which does not use ionising radiation. Liver
size is usually determined using single linear measurements rather
than an estimation of liver volume. Volume calculations of the liver
are more easily and rapidly undertaken from Computed Tomogra-
phy (CT) or Magnetic Resonance Imaging (MRI) scans and are
regarded as gold standard.
3
These imaging modalities are expen-
sive compared to ultrasound, are far less mobile and accessible and
have the additional disadvantage of radiation exposure associated
with CT
4
and claustrophobia associated with MRI.
5
Liver size can
also be assessed through clinical examination using a palpation/
percussion technique, however multiple studies have shown this
technique to be innacurate.
6e8
In the current Australian clinical environment the liver is
commonly measured during a routine upper abdominal ultra-
sound. Measurement of liver size using ultrasound is most often
determined by taking a simple linear measurement of the liver
from a plane along the mid-clavicular line, and using cut off values
to differentiate normal from abnormal liver size. Within the liter-
ature, two methods have been described using this plane:
measuring the maximum width between the anterior and posterior
borders of the liver and measuring the distance between the liver
dome to liver tip, with both methods using different upper limits to
differentiate normal from enlarged livers.
Niderau et al. (1983)
9
used the method measuring the
maximum width between the anterior and posterior borders of the
liver in 915 participants with normal livers to determine normal
liver size. They determined an upper limit cutoff value of 11.3 cm
for normal livers. Reliability and accuracy of this technique was not
reported and the upper limit cut off value was not tested for its
effectiveness in detecting enlarged livers.
Two studies used the liver dome to liver tip method.
10,11
Kratzer
et al. (2003)
11
determined 16 cm to be the upper limit cut off value
for normal livers from a sample of 2080 normal participants. Ac-
curacy and rater reliability for the measurement method was not
reported. Gosink & Leymaster (2005)
10
determined 15.5 cm as the
* Corresponding author. University of South Australia, City East Campus, BJ1-18
Frome Road, Adelaide 5000, Australia. Tel.: þ61 08 8302 2243.
E-mail addresses: Jessie.childs@unisa.edu.au (J.T. Childs), Kerry.Thoirs@unisa.
edu.au (K.A. Thoirs), Adrian.Esterman@unisa.edu.au (A.J. Esterman).
c
Tel.: þ61 8 8302 2903.
d
Tel.: þ61 8 8302 2163.
Contents lists available at ScienceDirect
Radiography
journal homepage: www.elsevier.com/locate/radi
http://dx.doi.org/10.1016/j.radi.2015.12.009
1078-8174/© 2016 The College of Radiographers. Published by Elsevier Ltd. All rights reserved.
Radiography 22 (2016) e125ee130