Radiography xxx (2016) 1e8

Contents lists available at ScienceDirect

Radiography
journal homepage: www.elsevier.com/locate/radi

Is magnetic resonance imaging a viable alternative to ultrasound as

the primary imaging modality in the diagnosis of paediatric
appendicitis? A systematic review
G. Ogunmefun a, *, M. Hardy b, S. Boynes b
a
Radiology Department University College Hospital Queen Elizabeth Road, Ibadan, Oyo State, Nigeria
b
Faculty of Health Studies, University of Bradford, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: Background: Appendicitis is the most common cause of acute abdominal pain requiring surgical inter-
Received 30 October 2015 vention in paediatric patients. Ultrasound is generally the diagnostic imaging modality of choice, fol-
Received in revised form lowed by CT, where paediatric appendicitis is suspected. However, high operator dependency and
10 December 2015
diagnostic restrictions related to anatomical and clinical presentation may limit consistency of appli-
Accepted 2 January 2016
Available online xxx
cation. This paper explores whether MRI is a viable alternative to ultrasound as the primary imaging
modality.
Method: A systematic review of the literature was undertaken. A search of Medline, Cinahl, PubMed
Keywords:
Child
Central and Google Scholar was undertaken supplemented by a review of reference lists, author
Paediatric searching and review of NICE evidence base for existing guidelines. Included studies were assessed for
Appendicitis bias using the QUADAS-2 quality assessment tool and data were extracted systematically using a pur-
Magnetic Resonance Imaging (MRI) posefully designed electronic data extraction proforma.
Results: Seven studies were included in final review. The age range of participants extended from 0 to 19
years. Only one study with a patient age range of 0e14 used sedation. Sensitivity estimates from the
included studies ranged from 92% to 100% while specificity ranged from 89% to 100%. A significant
variation in the number and type of sequences was noted between the studies.
Conclusion: MRI offers high sensitivity and specificity comparable to contrast enhanced CT and greater
than ultrasound as reported in the literature. Where accessibility is not a restriction, MRI is a viable
alternative to ultrasound in the assessment and diagnosis of paediatric appendicitis. Clinical practice
recommendations have been provided to facilitate the translation of evidence into practice.
© 2016 The College of Radiographers. Published by Elsevier Ltd. All rights reserved.

Introduction In the United States of America, the annual incidence of

Appendicitis is defined as the inflammation of the vermiform
appendix, often as a consequence of bacterial aggregation, resulting
in the appendix becoming inflamed and pus filled.1e3 Appendicitis
is the most common cause of acute abdominal pain requiring sur-
gical intervention in paediatric patients.4e6 The global incidence of
appendicitis in children increases from 1 to 2 cases per 10,000
children aged less than 4 years to 25 cases per 10,000 in children
between the ages of 10 and 17 years.7
* Corresponding author. Tel.: þ44 7557415188, þ234 8033739297.
E-mail address: graciousgbo65@yahoo.com (G. Ogunmefun).
appendicitis is 37 cases per 10,000 children aged between 0 and 14
years8 with approximately 70,000 cases reported among all
American children annually.9 In the United Kingdom, acute
appendicitis accounts for an estimated 34,000 hospital admissions
among the general population with 20% of these cases (approxi-
mately 6800) reported to be in children of 0e14 years of age.10
The diagnosis of acute appendicitis can be very challenging. In
adults, appendicitis often presents with a typical progression of
symptoms: periumbilical pain progressing to nausea, right lower
quadrant pain and eventually vomiting and fever.11,12 As a result,
successful diagnosis can often be made on presenting clinical fea-
tures and results of laboratory tests (e.g. Total Leukocyte Count
(TLC), C-reactive protein; neutrophil count).11,13e15 In children,
appendicitis may not present with such typical symptoms.11,12

http://dx.doi.org/10.1016/j.radi.2016.01.001
1078-8174/© 2016 The College of Radiographers. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ogunmefun G, et al., Is magnetic resonance imaging a viable alternative to ultrasound as the primary imaging
modality in the diagnosis of paediatric appendicitis? A systematic review, Radiography (2016), http://dx.doi.org/10.1016/j.radi.2016.01.001
2 G. Ogunmefun et al. / Radiography xxx (2016) 1e8
Instead, while childhood appendicitis may present initially with
periumbilical pain, symptom progression may lead to flatulence,
bowel irregularity/diarrhoea, indigestion, and general malaise.16
Consequently, a substantial proportion of paediatric appendicitis
cases may be misdiagnosed if clinical decision making is based on
physical examination, symptoms and laboratory investigations
alone.17 To prevent misdiagnosis and reduce negative appendec-
tomy rates, imaging has been recommended as part of the diag-
nostic pathway.16
Ultrasound (US) is generally the diagnostic imaging modality of
choice where paediatric appendicitis is suspected as it is: readily
available; has no radiation risk; is relatively fast; and, in compari-
son to other cross-sectional imaging modalities, is relatively inex-
pensive.18 High sensitivity (88%: 95%CI[86e90]) and specificity
(94%: 95%CI[94e96]) have also been documented for US in the
assessment of paediatric appendicitis.18,19 However, the focussed
nature of ultrasound assessment limits its contribution in deter-
mining alternative causes of presenting symptoms20 and operator
dependency remains an acknowledged fundamental limitation.18
Further, the accurate diagnosis of appendicitis using US may be
restricted as a consequence of bowel gas or distension, patient
obesity and a retro-caecal (deeply situated) appendix.21
Computed Tomography (CT) has previously been considered the
main alternative to US and has a high sensitivity and specificity in
the diagnosis of appendicitis with much reduced operator de-
pendency.21 A meta-analysis by Doria et al. (2006)19 compared CT
and US in the diagnosis of appendicitis and determined that in the
diagnosis of paediatric appendicitis, the evidence reviewed sug-
gested that CT had a pooled sensitivity of 94% (95%Cl: 92 to 97) and
a pooled specificity of 95% (95% CI: 94 to 97). However, CT also
presents a far greater risk of harm to the child from exposure to
ionising radiation and reaction to intravenous contrast media.22 As
a result, the trend is not to refer paediatric patients for CT where
appendicitis is suspected.22
Magnetic Resonance Imaging (MRI) has been successfully
adopted in the assessment of appendicitis in pregnant women of-
fering high soft tissue contrast without the use of radiation.20
However, long examination times and limited clinical availability
have been cited as major limitations to the wider application of MRI
in the assessment of acute clinical conditions requiring timely de-
cision making.20 In addition, long examination times and move-
ment restriction requirements previously meant that sedation of
children may have been considered necessary to reduce anxiety
and optimise MRI (and perhaps CT) imaging outcomes. Today, the
greater accessibility to MRI within the clinical radiology setting
globally, and the development of new and faster imaging se-
quences, reduces the impact of these concerns when identifying
MRI as the diagnostic imaging modality of choice. As a result, it is
time to consider whether MRI should be considered a viable
alternative to ultrasound as the primary imaging modality in the
assessment of paediatric appendicitis. This paper reports the find-
ings of a systematic review of the research evidence and considers
whether MRI should form part of the diagnostic pathway where
paediatric appendicitis is suspected and explores the optimal
diagnostic scan sequences to reduce examination time. No previ-
ously published systematic reviewed has explored the value of MRI
as the index test in the assessment of paediatric appendicitis and
therefore this review provides a significant contribution to the
evidence base.
Method
A search of Medline, Cinahl and PubMed central databases and
Google Scholar was undertaken supplemented by hand searching
of key imaging journals (e.g. British Journal of Radiology;
Please cite this article in press as: Ogunmefun G, et al., Is magnetic resonance imaging a viable alternative to ultrasound as the primary imaging
modality in the diagnosis of paediatric appendicitis? A systematic review, Radiography (2016), http://dx.doi.org/10.1016/j.radi.2016.01.001
Radiography), review of reference lists, author searching and re-
view of the NICE (National Institute for Health & Care Excellence)
evidence base for existing guidelines. Citations were identified
using the following key search terms (and their alternatives):
Magnetic Resonance Imaging (MRI; MR; Nuclear Magnetic Reso-
nance (NMR)); Appendicitis (vermiform appendix, epiphylitis).The
search was limited to primary research studies published from
January 2005 to April 2015 to take account of the recent advances in
MRI pulse sequences and the broader clinical application of the
technology.
Following the identification of all potentially relevant research
studies, the titles and abstracts were screened to determine
whether they met the inclusion/exclusion criteria. These criteria
were derived from the primary research question “is MRI a viable
alternative to Ultrasound as the primary imaging modality in the
diagnosis of paediatric appendicitis” and are listed in Table 1. De-
cision making around inclusion was based on the “rule out” prin-
ciple with papers only being rejected where the reviewer was
certain of their lack of relevance. At each stage, if uncertainty
existed over whether a paper should be included in the review, the
paper was retained. The full text of all retained articles was
examined to make the final decision on inclusion/exclusion.
All retained full text papers were independently assessed for
quality using the Quality Assessment of Diagnostic Accuracy
Studies (QUADAS-2)23 by at least 2 authors and data were extracted
directly into a Microsoft Excel24 spread sheet using a purposefully
pre-designed extraction framework to promote consistency. Any
inconsistencies in opinion re inclusion or paper quality were
resolved through discussion and consensus agreement. Paper
quality was documented using an adaptation of the QUADAS-2
assessment checklist summary criteria24 (Table 2). A summary
value was awarded to each paper in Table 3 with ‘High’ repre-
senting a study with low risk of bias and low concerns regarding
applicability of study findings, ‘Average’ representing a study with
an unclear risk of bias and unclear concerns regarding applicability,
and ‘Low’ representing a study with a high risk of bias and high
concerns regarding applicability. Only papers considered to be of
high or average quality were retained in the final evaluation
(Table 3).
Data analysis was by descriptive synthesis and comparison of
extracted data. Meta-analysis or pooling of extracted data was not
appropriate due to the diversity of study designs, variations in
clinical characteristics and technical parameters, and differences in
how diagnostic accuracy was determined.
Table 1
Inclusion/Exclusion criteria.
Question Inclusion Exclusion
facet
Population Children 0e19 years with Adults studies involving children
suspected appendicitis. but where child data could not be
independently evaluated.
Index test MR Studies not involving MRI or where
MRI data cannot be independently
evaluated.
Comparison Ultrasound Other imaging modalities
test
Outcome Sensitivity and specificity
measures Diagnostic accuracy
Examination time
Study Diagnostic test studies Studies without comparison
design between ultrasound and MRI.
Qualitative studies
Reference Surgical confirmation
standard Histopathology
Clinical follow-up of
symptoms
 G. Ogunmefun et al. / Radiography xxx (2016) 1e8 3

Table 2
QUADAS-2 appraisal of included studies.

Study Risk of bias Applicability concerns

Patient selection Index test Reference standard Flow and timing Patient selection Index test Reference standard
25
Bayraktutan et al. (2014) Low Low Low Unclear Low Low Low
Herliczek et al.26 (2013) Low Low Low Low Low Low Low
Johnson et al.27 (2012) Unclear Low Low Low Low Low Low
Moore et al.28 (2012) Low Low Low Low Low Low Low
Orth et al.29 (2014) Unclear Low Low Unclear Low Low Low
Rosines et al.30 (2014) High Low Low Unclear Low Low Low
Thieme et al.31 (2014) Low Low Low Low Low Low Low

Results studies utilized at least two or more T2 weighted sequences. With

A flow chart detailing the review process is provided in Fig. 1.
Details of the included studies are summarized in Table 3 with
quality assessment grading identified in the final column. Details of
the index test characteristics and extracted outcome measures are
detailed in Tables 4 and 5 respectively.
Sensitivity estimates
All the included studies used pathology following surgery, or
clinical follow-up of those patients receiving conservative care, as
the reference standard. Sensitivity across the included studies
ranged from 92% to 100%. Only one study (Rosines et al.) employed
Gadolinium contrast agent (Gadobenate Dimeglumine (Bracco di-
agnostics) administered intravenously with a weight based dosage
of 0.2 ml per kg) and they reported an examination sensitivity of
94%, which is lower than the sensitivities reported by Herliczek
et al., Johnson et al., Moore et al., and Thieme et al. The study by
Bayraktutan et al. explicitly set out to evaluate the accuracy of
Diffusion Weighted Imaging (DWI) and reported a sensitivity of
78%, 81%, and 92% for DWI, conventional MRI sequences, and
combined DWI and conventional sequences respectively.
Specificity estimates
Specificity across the included studies ranged from 89% to 100%.
The study by Rosines et al., employing Gadolinium contrast agent,
reported a specificity of 100% as did the study by Bayraktutan et al.
The study by Thieme et al. reported the lowest specificity at 89%
with all other studies reporting sensitivities of 96% or higher. The
study by Bayraktutan et al. evaluating the accuracy of DWI reported
a specificity of 67%, 100% and 100% for DWI, conventional MRI se-
quences, and combined DWI and conventional sequences
respectively.
Variations in diagnostic sequence pathway and scan time
Scan time is dependent on several variables including magnetic
field strength, RF coil type and pulse sequences used. All studies,
with the exception of Herliczek et al., used a multichannel phased
array coil. Herliczek et al. did not indicate the type of coil used. The
studies by Johnson et al. and Herliczek et al. utilised a 3T
magnet although Herliczek et al. combined this with a 1.5 T magnet.
All other studies used 1.5T magnets which remains the most
common magnet field strength in the UK. With the exception of
Herliczek et al., all studies standardised their within study imaging
sequences but there was significant variation between the studies
with respect to the sequences employed. Four studies, Bayraktutan
et al., Herliczek et al., Orth et al., and Rosines et al. included T1
sequences together with T2 weighted images. However, contrast
enhanced T1 sequences were only employed by Rosines et al. All
the exception of Bayraktutan et al. and Orth et al., single shot se-
quences featured heavily in the studies. Balanced gradient echo
sequences were used by Rosines et al. and Thieme et al. Diffusion
Weighted Imaging (DWI) was employed in three studies (Bayr-
aktutan et al., Orth et al. and Thieme et al.). Fat saturation was used
by all studies except Rosines et al. and Thieme et al. and a volume
3D T2-weighted FSE SPACE, a volumetric isotropic acquisition, was
utilized by Herliczek et al.
Given the variation in field strength and sequences used it is
not surprising that scan time varied among the included studies.
Four studies (Johnson et al., Moore et al., Orth et al. and Thieme
et al.) recorded scan times of less than 20 min per patient.
Johnson et al. reported a median scan time of 5 min 40 s with all
examinations completed within 8 min and 45 s. This study was
performed using a 3T magnet and four fast T2 weighted se-
quences. The longest time was reported by Herliczek et al. with a
mean image acquisition time of 30.5 min but with a reported time
range of 10e66 min, this may be influenced by a small number of
lengthy examinations. No median time was reported to confirm
this. Bayraktutan et al. and Rosines et al. did not provide exact
data on scan time.
Discussion
The patient pathway, modality and techniques assessment of
acute appendicitis in children has been the subject of debate
internationally.32 Contemporary pre-operative imaging modalities
in the evaluation of paediatric appendicitis are currently ultrasound
(US) as the primary tool and CT as the secondary complimentary
imaging modality33 as recommended by the Royal College of Ra-
diologists in the UK.34 An ionising radiation-free cross sectional
imaging pathway of US, selectively followed by MRI where US
findings are inconclusive, is a new and novel approach to the
diagnosis of paediatric appendicitis that is being considered by the
imaging community.35 However, with the acknowledged limita-
tions of ultrasound and the greater accessibility to faster MRI
technologies, it is likely that questions will soon be asked around
preferred primary imaging approach. This systematic review is the
first to consider whether MRI is a viable alternative to US as the
primary imaging modality in the diagnosis of paediatric
appendicitis.
Seven primary research studies met the inclusion criteria for
this review. Reported sensitivities for MRI were high across all
included studies ranging from 92% to 100% with three studies
reporting 100% sensitivity. Reported specificities were also high
across included studies ranging from 89% to 100% with two studies
reporting 100% specificity. Reported positive (PPV) and negative
(NPV) predictive values were also relatively high across the studies,
despite the varying prevalence of appendicitis within the cohorts
examined, suggesting that the diagnostic outcome of MRI exami-
nation is accurate and reliable.

Please cite this article in press as: Ogunmefun G, et al., Is magnetic resonance imaging a viable alternative to ultrasound as the primary imaging
modality in the diagnosis of paediatric appendicitis? A systematic review, Radiography (2016), http://dx.doi.org/10.1016/j.radi.2016.01.001
4 G. Ogunmefun et al. / Radiography xxx (2016) 1e8

Table 3
Characteristics and quality assessment (QA) of included studies.

Study Title Objectives Country Study type Number Mean age Positive diagnostic criteria QA
author/Year of origin (n) (age range)

Bayraktutan Diagnostic Performance of
et al.25 Diffusion-Weighted MR
(2014) Imaging in Detecting Acute
Appendicitis in Children:
Comparison with Convectional
MRI and Surgical Findings
Herliczek Utility of MRI After
et al.26 Inconclusive Ultrasound in
(2013) Paediatric Patients with
Suspected Appendicitis:
Retrospective Review of 60
Consecutive Patients.
Johnson Ultrafast 3T MRI in the
et al.27 Evaluation of Children With
(2012) Acute Lower Abdominal Pain
for the Detection of
Appendicitis
Moore MRI for clinically suspected
et al.28 paediatric appendicitis: an
(2012) implemented program
Orth et al.29 Prospective Comparison of
(2014) MR Imaging and US for the
Diagnosis of Paediatric
Appendicitis
Rosines Value of Gadolinium-Enhanced
et al.30 MRI in Detection of Acute
(2014) Appendicitis in Children and
Adolescents
Thieme Diagnostic accuracy and
et al.31 patient acceptance of MRI in
(2014) children with suspected
appendicitis
To determine the value of
diffusion-weighted MRI in
diagnosis of acute appendicitis
in Children
To examine the utility of
appendix MRI in evaluation of
paediatric patients with right
lower quadrant pain and
inconclusive sonography
appendix sonography findings
Feasibility of ultrafast 3T MRI
in the evaluation of children
with acute lower abdominal
pain for the detection of
appendicitis
To describe the development
and results of a fully
implemented clinical
programme using MRI as the
primary imaging modality.
To prospectively compare non- USA
enhanced MR imaging and
Ultrasound for the diagnosis of
paediatric appendicitis
To determine both the value
of gadolinium-enhanced MRI in
children with suspected acute
appendicitis and the best
sequences for detecting acute
appendicitis, to thereby
decrease imaging time
To compare Magnetic
Resonance Imaging (MRI) and
Ultrasound (USS) in children
with suspected appendicitis.
Turkey Prospective. 45 7 years Thickening of the Average
Cohort study. (0e14) appendix > 6 mm in outer
Single centre. diameter, intraluminal fluid,
and inflammation of the
periappendeceal tissue.
Periappendiceal fluid collections
and markedly hyperintensity on
DWI, hypointensity on ADC maps
U.S.A Retrospective 60 13.4 years Appendix visualization, Average
Cohort study. (7e17) appendiceal diameter (>7 mm
Single centre. was considered to be pathologic),
appendeceal mural thickening of
more than 3 mm, appendicieal
mural edema, the presence of
appendiceal intraluminal fluid.
The presence of appendiceal
inflammation, the presence of
RLQ inflammation and the
presence of an appendicolith
U.S.A Prospective 42 11.4 years Markedly hyperintense Average
Cohort study. (4e17) thickened wall, marked
Single centre. hyperintense periappendiceal
tissue, and dilated appendix
(6 mm) which can often be
accompanied by free fluid in
the pelvis.
USA Retrospective 208 11.2 years Focal periappendiceal High
Cohort study. (3e17) inflammation and fluid.
Single centre Appendeceal enlargement,
intraluminal high signal intensity
and signal void consistent with
appendicolith
Prospective 81 12.4 years Appendiceal dilatation, wall Average
Cohort study. (4e17) thickening or inflammation,
Single centre presence of an appendicolith,
periappendiceal inflammation as
indicated by increased signal
intensities as indicted by fluid
sensitive sequences and or
supraphysiologic free fluid or a
loculated fluid collection.
USA Retrospective 49 12.9 years visualization of an enlarged Average
Cohort study. (7e19) appendix greater than 7 mm in
Single centre diameter, periapendical fat
stranding; mucosal
enhancement for contrast-
enhanced sequences
Holland Prospective 104 12 years Appendix diameter, presence of High
Cohort study. (4e18) peri-appendiceal infiltration,
Single centre fluid, abscess, appendicolith,
lymphadenopathy and restricted
diffusion

A diagnosis of appendicitis was based on defined criteria and all
studies considered an enlarged outer diameter of the appendix to
be an important criterion. However, the numerical measure used to
determine an enlarged diameter varied slightly across included
studies with Bayraktutan et al. and Johnson et al. fixing their
diagnostic boundary at  6 mm while Rosines et al. and Herliczek
et al. fixed the diameter at  7 mm. The remaining three studies did
not define the numerical value for the diameter but instead
mentioned a thickened appendiceal wall, presumably assessed
subjectively on image review. With the diagnostic boundary set at
 6 mm, Bayraktutan et al.25 and Johnson et al.27 reported sensi-
tivities of 92%25 and 100%27 and specificities of 100%25 and 99%27
respectively. With the diagnostic boundary set at  7 mm, Ros-
ines et al.30 and Herliczek et al.26 reported sensitivities of 94%30 and
100%26 and specificities of 100%,30 96%26 respectively. It can be
therefore be inferred that the 1 mm difference in appendiceal outer
diameter measurement made very little difference to reported
sensitivity and specificity values and as such, the smaller value
(6 mm) is recommended to reduce risk of false negative findings.
Field strengths of 1.5T and 3.0T were employed by included
studies. Johnson et al. utilised a field strength of 3.0T and reported a
sensitivity of 100% and specificity of 99%. Herliczek et al. used a
combination of 1.5T and 3.0T field strengths and reported a sensi-
tivity of 100% and 96% but distinction was not made as to any
variation in sensitivity or specificity between the 2 field strengths.
At 3.0T, there is an increase in Signal to Noise Ratio (SNR) and
Contrast to Noise Ratio (CNR) which may lead to improved image
resolution and shortened scan times.36 These gains in SNR are more
pronounced in T2-weighted images than T1 sequences.36 However,
at 3.0T there is potentially greater magnetic susceptibility and

Please cite this article in press as: Ogunmefun G, et al., Is magnetic resonance imaging a viable alternative to ultrasound as the primary imaging
modality in the diagnosis of paediatric appendicitis? A systematic review, Radiography (2016), http://dx.doi.org/10.1016/j.radi.2016.01.001
 G. Ogunmefun et al. / Radiography xxx (2016) 1e8 5

AddiƟonal records through web

Database searching searches; hand searches, author and
n=138 citaƟon searches
n=72

Duplicates
n =70

Remaining records remained aŌer duplicates removed

n=140

Studies excluded at
Ɵtles/abstracts
n=100

Full text arƟcles reviewed

n=40
Studies excluded at full text
review (Literature review, case
study; unrelated outcome
measures; data not available)
n=33

Studies included in the review

n=7

Figure 1. Flow diagram of included studies.

chemical shift artefact with radio frequency pulse inhomogeneity
also being a potential source of artefact in large or obese children
and young people.36 Due to the limited data available within the
included studies in relation to field strength variations, it is not
possible to determine the added value of 3.0T MRI imaging over
and above 1.5T MRI with respect to diagnostic accuracy in paedi-
atric appendicitis and further work is required to explore this.
The use of Gadolinium in the examination of appendicitis has
been advocated, as enhanced T1 weighted sequences are useful for
subtle mucosal enhancement and establishing complications with
appendicitis.37 However, only Rosines et al. employed Gadolinium
contrast agent within this review and as their reported sensitivity
and specificity values were comparable to those reported by other
authors, it appears unlikely that Gadolinium is a requirement for
optimising the diagnosis of paediatric appendicitis. This suggestion
is supported by the finding that non-contrast enhanced MRI is
equivalent to contrast enhanced CT in the assessment of the ap-
pendix in pregnant women.38 However, given the wide variation in
technical parameters between studies included in this review,
further work is required to confirm the contribution of Gadolinium
contrast media within the context of paediatric appendicitis before
recommendations for its adoption or exclusion can be made with
confidence.
A diverse range of pulse sequences were used within the
included studies. Bayraktutan et al. evaluated the improvement in
diagnostic accuracy as a result of including DWI as an adjunct to
conventional sequences. DWI demonstrates an inflamed appen-
dix and surrounding fat as high signal, secondary to a restriction
in water diffusion.39 DWI has previously been shown to have high
sensitivity and specificity values and therefore useful in
discriminating between perforated and non-perforated
appendicitis.40,41
T2-weighted sequences were used by all studies and, when
combined with ultra-fast sequences, are noted to greatly improve
assessment of lesions in the small bowel.42 T2 weighted sequences
are very sensitive to fluid filled pathology such as oedema which
appears as a hyper-intense signal.36,43 The T2 FSE 3D (SPACE)
sequence utilized by Herliczek et al. is an approach to volumetric
acquisition that allows retrospective reformatting in multiple ori-
entations of the appendix.44 Fat saturated sequences demonstrate
oedema or inflammatory processes as a hyper-intense signal within
the wall or adjacent to fat.43,45
T1 weighted sequences utilised by four studies in this review
can be useful in demonstrating the normal appendix.36,45 T1
weighted GRE sequences were employed by Orth et al. and Rosines
et al. and may be useful in the diagnostic work up for acute
appendicitis as gas or negative oral contrast within the appendix
(suggesting luminal patency) will demonstrate susceptibility arte-
fact on these sequences.46 The unenhanced T1 weighted sequence
has been shown to be of little use in demonstrating an inflamed
appendix. Evidence from the diagnostic outcome measures of the
primary articles within this review supports this view as the
diagnostic outcomes of the three studies which did not employ any
T1 weighted sequence (Johnson et al., Moore et al., Thiemme et al.)
were comparable to those studies that did. Enhanced T1 weighted
sequences add to the imaging time, thereby raising the risk of
perforation, and also increase the risk of Nephrogenic Systemic
Fibrosis (NSF) in patients with impaired renal function.47 Johnson
et al., Thieme et al. and Bayraktutan et al. also made use of parallel
imaging to reduce scan time. Parallel imaging enables faster image

Please cite this article in press as: Ogunmefun G, et al., Is magnetic resonance imaging a viable alternative to ultrasound as the primary imaging
modality in the diagnosis of paediatric appendicitis? A systematic review, Radiography (2016), http://dx.doi.org/10.1016/j.radi.2016.01.001
6 G. Ogunmefun et al. / Radiography xxx (2016) 1e8

Table 4
Index test characteristics extracted from included studies.

Magnet Coil type Number of Pulse sequence Scan time Gadolinium Sedation
strength sequences agent

Bayraktutan et al.25 1.5T 18 channel phased 7 FSE T1 axial Not Disclosed None 26 (58%) of patients
(2014) array FSE T1 coronal were sedated
FSE T2 axial STIR
FSE T2 coronal STIR
DWI (b50, b400 and b800)
Herliczek et al.26 1.5T (49 Not Disclosed 5 Protocol varied but sequences 10e66 min; mean None None
(2013) patients) included: time 30.5 min
3.0T (11 STIR
patients) 3D FSE T2 coronal SSFSE T2
(HASTE)
True FISP axial
In and out of phase T1
Johnson et al.27 3.0T 16 channel 4 SS-TSE T2 axial 2 min 29 se8 min None None
(2012) SS-TSE T2 coronal 45 s. Median time:
SS-TSE T2 sagittal 5 min 40 s.
SS-TSE T2 axial STIR
Moore et al.28 1.5T 4 or 6 channel 4 SS-TSE T2 coronal Median time: None None
(2012) phased array SS-TSE T2 axial 12 min, Mean time:
SS-TSE T2 axial STIR 14.2 min
Coronal SPAIR
Orth et al.29 (2014) 1.5T Multichannel 6 FSE T2 coronal 17 min None None
Coil FSE T2 axial
FSE T2 axial STIR
GRE T1
DWI b50 and b400
Rosines et al.30 1.5T 8 channel 6 Balanced SSFP T1 axial Not stated IV Contrast None
(2014) Balanced SSFP T1 coronal (Suggests Scan time (Multihance)
Balanced SSFP T2 axial can be reduced to
Balanced SSFP T2 coronal 10e15 min if
Contrast enhanced balanced balanced SSFP
SSFP T1 axial sequences
Contrast enhanced balanced excluded)
SSFP T1 coronal
Thieme et al.31 1.5T Phased array 6 True FISP (axial and coronal) 12 min in room None None
(2014) Haste (axial and coronal) time
DWI (axial and coronal)

FSE¼ Fast Spin Echo; STIR¼ Short T1 Inversion Recovery; DWI ¼ Diffusion Weighted Imaging; EPI ¼ Echo Planar Imaging; SSFSE¼ Single shot Fast Spin Echo; TSE ¼ Turbo Spin
Echo; FISP¼ Coherent gradient echo; HASTE¼ Half-Fourier Acquisition Single-Shot Turbo Spin-Echo; SSFP¼ Steady State Free Procession; GRE ¼ GRadient Echo; SPAIR¼
Spectrally Attenuated Inversion Recovery.

Table 5
Extracted outcome measure data.

Bayraktutan et al. Herliczek et al. Johnson et al. Moore et al. Orth et al. Rosines et al. Thieme et al. (2014)31
(2014)25 (2013)26 (2012)27 (2012)28 (2014)29 (2014)30

True Positive (TP) 33 10 Not Discernable 40 29 Not Disclosed 58

False Positive (FP) 0 2 Not Discernable 5 1 Not Disclosed 5
True Negative (TN) 3 48 Not Discernable 162 50 Not Disclosed 41
False Negative (FN) 3 0 Not Discernable 1 1 Not Disclosed 0
Sensitivity 92% 100% 100% 97.6% 93.3% 94% 100%
Specificity 100% 96% 99% 97% 98% 100% 89%
95% Confidence Sensitivity: Sensitivity: Sensitivity: Sensitivity: Sensitivity: Not Disclosed Sensitivity: 0.92e1.00
Interval 0.775e0.983 Specificity: 0.72e1.00 0.93e1.00 0.871e0.999 0.779e0.992 Specificity: 0.77e0.95
0.292e1.00 Specificity: 0.87 Specificity: Specificity: 0.932 Specificity:
(calculated from data) e0.98 0.95e1.00 e0.99 0.896.-100
Positive Predictive 100% 83% 98% 88.9% 96.5% 100% 92%
Value (PPV)
Negative Predictive 50% 100% 100% 99.4% 96.2% 97% 100%
Value (NPV)

acquisition resulting in fewer motion artefacts and improving res-
olution.48 The remaining studies did not make reference to the use
of parallel imaging and therefore its contribution to diagnosis and
examination speed are unclear.
This systematic review has focused on the ability of MRI to
enable accurate diagnosis of acute paediatric appendicitis. Acute
Appendicitis is a progressive clinical condition occurring over a
period of 24e36 h.49 The number of sequences employed greatly
influences the scan time per patient. By limiting the number of
sequences acquired to four, as well as employing T2 ultra-fast se-
quences, parallel imaging and 3T field strength, Johnson et al. were
able to report the fastest scan times without evidence of compro-
mising diagnostic accuracy compared to other studies in this re-
view. While an abdominal US examination typically takes
20e30 min in children,50 US examination for the evaluation of
paediatric appendicitis may take even longer as a child may

Please cite this article in press as: Ogunmefun G, et al., Is magnetic resonance imaging a viable alternative to ultrasound as the primary imaging
modality in the diagnosis of paediatric appendicitis? A systematic review, Radiography (2016), http://dx.doi.org/10.1016/j.radi.2016.01.001
 G. Ogunmefun et al. / Radiography xxx (2016) 1e8
become uncooperative and experience pain with the rebound
technique at McBurney's point.50 The findings of this review sug-
gest that if the number of MRI sequences can be restricted and
tailored to T2 fluid filled and fat suppressing sequences with a
phase array coil, then it is possible that MRI examination time for
assessment of the paediatric appendix may be shorter than an US
examination without compromising accuracy. This focussed
approach to MRI assessment of the appendix may reduce its
contribution to determining differential diagnoses (e.g. Crohn's
disease, endometriosis, Meckel's diverticulum)51 but this argument
could also be made for other focussed examinations across imaging
modalities specifically directed to answering the clinical question
posed.
A previous restriction on the use of MRI for acute paediatric
conditions has been the increased use in sedation to reduce anxiety
and improve compliance with movement restrictions. In this re-
view, only Bayraktutan et al. sedated patients with 58% of partici-
pants receiving sedation. This may be as a consequence of the wide
age range of participants (0e14 years) as the remaining included
studies restricted the lower age range to 3 years minimum and
permitted parents/guardians to accompany the child into the
magnet room. Given the comparability in diagnostic outcomes
across studies reviewed, it would appear that sedation is not
required in the assessment of acute appendicitis in children aged 3
years and over and should no longer be a perceived barrier to
access.
Limitations/strength
This systematic review was undertaken as part of a Master of
Science award at the University of Bradford. The initial review of
literature and evaluation was completed independently by GO
under the supervision of MH. For purpose of publication, the review
process was repeated with independent evaluation by SB and
mediation by MH to ensure rigour of systematic review process.
Inclusion was limited to papers published between 2005 and April
2015 to accommodate innovations in MRI technology and wider
application within clinical practise. Heterogeneity was noted across
the included studies which prevented meta-analysis of the data.
Despite this limitation, analysis of the data has suggested that MRI
may well be a viable alternative to ultrasound in the diagnosis of
acute appendicitis in children.
Practice recommendation
In order to facilitate the application of evidence into practice,
the evidence from this review has been summarised into the series
of recommendations below. Further evaluation of the application of
these recommendations in practice is required for validation.
 Where MRI is available 24 h per day, the evidence from this
review suggests that MRI may be designated as the primary
imaging tool to investigate suspected paediatric acute appen-
dicitis as the diagnostic accuracy (sensitivity and specificity) is
comparable to those reported for CT and greater than those
reported for Ultrasound.
 The use of contrast agents is not necessary for the diagnosis of
appendicitis in children.
 The use of sedation is not required routinely in children aged 3
years or over and, where possible, parental attendance in the
magnet room is to be encouraged to reduce child anxiety and
promote compliance.
 Where the outer diameter of the appendix is used as a diag-
nostic decision making boundary, a measure of  6 mm should
be applied.
Please cite this article in press as: Ogunmefun G, et al., Is magnetic resonance imaging a viable alternative to ultrasound as the primary imaging
modality in the diagnosis of paediatric appendicitis? A systematic review, Radiography (2016), http://dx.doi.org/10.1016/j.radi.2016.01.001
7
 A focussed sequence pathway using a phased array coil and
parallel imaging is recommended to optimise diagnosis and
minimise scan time. From the evidence reviewed, the optimum
sequence selection is:
 FSE/SSFSE T2 axial
 FSE/SSFSE T2 coronal
 FSE/SSFSE T2 sagittal
 FSE/SSFSE T2 axial fat saturation
 FSE/SSFSE DWI (axial and coronal b500)
Conclusion
The evidence from this review suggests that MRI has high
sensitivity and specificity in relation to the accurate diagnosis of
acute appendicitis in children, comparable to those reported for
contrast enhanced CT and greater than Ultrasound. Consequently,
we conclude that is a viable alternative to ultrasound and may be
adopted as the primary imaging modality of choice, where acces-
sibility is not restricted, or used to complement ultrasound where
findings are indeterminate or inconclusive. Practice recommenda-
tions have been provided based upon the evidence reviewed and
these require validation within the practice setting.
Conflict of interest
The authors have no conflict of interest to disclose
References
1. Humes DJ, Simpson J. Acute Appendicitis. BMJ 2006;333(2):530e4.
2. Minneci PC, Sulkowski JP, Nacion KM, Mahida JB, Cooper JN, Moss RL, et al.
Feasibility of non-operative management strategy for uncomplicated acute
appendicitis in children. JACS 2014;219(2):272e9.
3. Dorland’s illustrated Medical Dictionary. 26th ed. Philadelphia: W. B. Saunders
Company; 1981.
4. Flum DR, Koepsell T. The clinical and economic correlates of misdiagnosed
appendicitis: nationwide analysis. Arch Surg 2002;137(7):799e804.
5. Eugenia C, Pavel C, Yehiel B. Sonography of acute appendicitis in a 9-month old
infant. J Ultrasound Med 2004;23(1):865e7.
6. Gardikis S, Giatromanolaki A, Kambouri K, Tripsianis G, Sivridis E, Vaos G. Acute
appendicitis in Preschoolers: a study of two different populations of children.
Italian J Paediatr 2011;37(35):301e7.
7. Rothrock SG, Pagane J. Appendicitis in children: emergency department diag-
nosis and management. Ann Emerg Med 2000;36(1):39e51.
8. Bellollio MF, O’neil A, Lam SHF. Paediatric appendicitis. 2014. http://ahcmedia.
com/articles/21693-pediatric-appendicitis (accessed 3rd October 2015).
9. Danielle W. Paediatric Appendicitis. SAGES. 2015. http://sages.org/wiki/
pediatric-appendicitis (accessed 10th October 2015).
10. Shah AP, Schnatterbeok P, Colin M. Appendicitis in children: is there a role for
routine ultrasonography. WLMJ 2010;2(3):37e47.
11. Sung T, Callahan MJ, Taylor GA. Clinical and imaging mimickers of acute
appendicitis in the paediatric population. AJR 2006;186(3):67e74.
12. Becker T, Khanbanda A, Bachur R. Atypical clinical features of paediatric
appendicitis. Acad Emerg Med 2007;14(2):124e9.
13. Van- Dieijen-Visser MP, GO PMNYH Brombacher PJ. The value of laboratory
test in patients suspected of acute appendicitis. Eur J Clin Chem Clin Biochem
1991;29(11):749e52.
14. Kamran H, Naveed J, Nazir A, Hameed M, Ahmed M, Khan U. Role of total
leukocyte count in diagnosis of acute appendicitis. J Ayub Med Coll Abbottabad
2008;20(3):70e1.
15. Tucker A, Sloan K, Gartsin I, Verghis R. White Cell Count, crp and appendicitis:
is there a role for pre-operative blood tests? A Cohort Study. J Health Med Inf
2015;6(2).
16. Nsutti R, Kruger D, Luvhergo TE. Clinical presentation of acute appendicitis in
adults at the chris hans Baragwanatt academic hospital. IJEM 2014;7(12):
353e61.
17. Mittal MK, Dayan PS, Macias CG, Bachur RG, Bannet J, Dudley NC, et al. Per-
formance of Ultrasound in the Diagnosis of Appendicitis in Children in a
Multicenter Cohort. Acad Emerg Med 2013;20(3):697e702.
18. Quigley AJ, Strafrace S. Ultrasound assessment of acute appendicitis: meth-
odology and pictorial overview of findings seen. Insights Imaging 2013;4(6):
742e51.
19. Doria AS, Moineddin R, Kellenbenger CJ, Epelman M, Beyene J, Schuh S, et al. US
or CT for diagnosis of appendicitis in children and adults? a meta-analysis.
Radiology 2006;241(1):83e94.
8 G. Ogunmefun et al. / Radiography xxx (2016) 1e8
20. Gaitini D. Imaging acute appendicitis: state of the art. J Clin Imaging Sci
2011;1(49):53e9.
21. Strouse PJ. Paediatric appendicitis: an argument for USS. Radiology
2010;255(1):8e13.
22. Leite NP, Perreira JM, Cunha R, Pinto P, Sirlin C. CT evaluation of appendicitis
and its complications: imaging techniques and key diagnostic findings. AJR
2005;185(2):406e17.
23. University of Bristol. QUADAS-2. 2015. http://www.bris.ac.uk/quadas/quadas-2/
(accessed 10th October 2015).
24. Brown S, Hutton B, Clifford T, Doug C, Grima D, Wells G, et al. A Microsoft based
tool for running and critically appraising network Meta-analysis-an overview
and application of NetMetaXL. Syst Rev 2014;3(110).
25. Bayraktutan U, Oral A, Kantarci M, Demir M, Ogul H, Yalcin A, et al. Diagnostic
performance of diffusion-weighted MR imaging in detecting acute appendicitis
in children: comparison with convectional MRI and surgical findings. J Magn
Reson Imaging 2014;39(3):1518e24.
26. Herliczek TW, Swenson DW, Mayo-Smith WW. Utility of MRI after inconclusive
ultrasound in paediatric patients with suspected appendicitis: retrospective
review of 60 consecutive patients. AJR 2013;200(5):969e73.
27. Johnson AK, Fillipi CG, Andrews T, Higgini T, Tam J, Keating D, et al. Ultrafast 3-
T MRI in the evaluation of children with acute lower abdominal pain for the
detection of appendicitis. AJR 2012;198(6):1424e30.
28. Moore MM, Gustas CN, Choudary AK, Methratta ST, Hulse MA, Geeting G, et al.
MRI for clinically suspected paediatric Appendicitis: an implemented program.
Paediatr Radiol 2012;42(9):1056e63.
29. Orth RC, Guillermann RP, Zhang W, Massand P, Bissett GS. Prospective com-
parison of MR imaging and US for the diagnosis of paediatric appendicitis.
Radiology 2014;272(1):233e40.
30. Rosines LA, Chow DS, Lampl BS, Chen S, Gordon S, Mui LW, et al. Value of
gadolinium enhanced MRI in detection of acute appendicitis in children and
adolescents. AJR 2014;203(5):543e8.
31. Thieme ME, Leeuwenburgh MMN, Valdehueza ZD, Bourman DE, De-Brun IGM,
Schreurs WH, et al. of MRI in children with suspected Appendicitis. Eur Radiol
2014;133(4):630e7.
32. Frush DP, Frush KS, Oldham KT. Imaging of acute appendicitis in children. EU
versus US or US versus CT? a North american perspective. Paediatr Radiol
2009;39(5):500e5.
33. Saito JM, Yan Y, Evashwick TW, Warner BW, Tarr PI. Use and accuracy of
diagnostic imaging by hospital type in paediatric appendicitis. Paediatrics
2012;131(1).
34. Royal College of Radiologists. iRefer. Making the best use of clinical radiology. 7th
ed. UK: Royal college of Radiologists; 2011.
Please cite this article in press as: Ogunmefun G, et al., Is magnetic resonance imaging a viable alternative to ultrasound as the primary imaging
modality in the diagnosis of paediatric appendicitis? A systematic review, Radiography (2016), http://dx.doi.org/10.1016/j.radi.2016.01.001
35. Aspelund G, Fingeret A, Gross E, Kessler D, Keung C, Thirumworths A, et al.
Ultrasonography/MRI versus CT for diagnosing appendicitis. Paediatrics
2014;133(4):586e93.
36. Chang KJ, Kamal IR, Macura KJ, Bluemke DA. 3.0-T MR imaging of the
Abdomen: comparison with 1.5T. Radiographics 2008;28(7):1993e8.
37. Baert AL, Reifer MF, Hricak H, Knauth M. MRI of the gastrointestinal Tract.
Illustrated ed.. Berlin: Springer Verlag; 2012.
38. Gore RM, Levine MS. Textbook of gastrointestinal radiology. 4th ed. Philadelphia:
Elsevier Saunders; 2015.
39. Luna A, Soto JA, Ribes R. Diffusion MRI outside the Brain. A case based review and
clinical applications. 2nd ed. Berlin: Springer-Verlag; 2012.
40. Avcu S, Cetin FA, Arslan H, Kemik O, Dulger AC. The value of diffusion-weighted
imaging and apparent diffusion coefficient quantification in the diagnosis of perfo-
rated and non-perforated appendicitis. Diagn Interv Radiol 2013;19(1):106e10.
41. Inci E, Kilickesmez O, Hocaoglu E, Bayramoglu S, Gmilli T. Utility of diffusion-
weighted imaging in the diagnosis of acute appendicitis. Eur Radiol
2011;21(4):768e75.
42. Reiser MF, Semmler W, Hrica KH. Magnetic resonance tomography. 3rd ed.
Berlin: Springer-Verlag; 2007.
43. LeBedis CA, Penn DA, Gupta A, Tkacz JN, Soto JA, Broder JC. Current application
of MRI in emergent gastrointestinal diseases. Appl Radiol J Pract Med Imaging
Manag 2012. http://www.appliedradiology.com/articles/current-applications-
of-mri-in-emergent-gastrointestinal-diseases (accessed 11th October 2015).
44. Siemens Space. Siemens healthcare GmbH. 2015. www.healthcare.siemens.
com/magnetic-resonance-imaging/options-and-upgrades/clinical-applications/
syngo-space (accessed 12th October 2015).
45. Cobben LPJ. Magnetic resonance imaging in acute appendicitis [Phd Thesis].
University of Amsterdam; 2009. www.dare.uva.nl/document/2/65624
(accessed 12th October 2015).
46. Grand DJ, Mayo-Smith WW, Woodfield C. A Practical body MRI: Protocol
application and image interpretations. 1st ed. Cambridge: Cambridge University
Press; 2012.
47. Kaewlai R, Abujudeh H. Nephrogenic systemic fibrosis. AJR 2012;199(1):17e23.
48. Deshmane A, Gulvani V, Griswold MA, Seiberlich N. Parallel MR imaging.
J Magn Imaging 2012;36(1):55e72.
49. Brennan GDG. Paediatric appendicitis: pathophysiology and appropriate use of
diagnostic imaging. CJEM 2006;8(6):425e32.
50. Cain T. Children (Paediatric) abdominal ultrasound inside radiology. 2009.
http://www.insideradiology.com.au/pages/view.php?T_id¼29#.VVC4Hv/vik
(accessed 12th October).
51. Levy AD, Mortele KJ, Yeh BM. Gastrointestinal Imaging. Illustrated ed.. Oxford:
Oxford University Press; 2015.