PICTORIAL REVIEW
Imaging in pulsatile tinnitu
G. Madani
a,
*
, S.E.J. Connor
b
, U
uly
are
al
agi
ate
iog
evie
audiological findings, the diagnostic yield of
Clinical Radiology (2009) 64, 319e328doi:10.1016/j.crad.2008.08.014Radiological investigation
If a mass is seen at otoscopy, then a thin-section
computed tomography (CT) of the petrous tempo-
ral bones is needed to assess the middle ear.
* Guarantor and correspondent: G. Madani, Radiology Depart-
ment, St Mary’s Hospital, Imperial College NHS trust, Praed
Street, London W2 1NY, UK. Tel.: þ44-2078861116; fax: þ44-
2078872281.
E-mail address: gittamadani@yahoo.com (G. Madani).
0009-9260/$ - see front matter ª 2008 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.such as migraine.
Radiological investigation aims to find treatable
causes of tinnitus. In the setting of non-PT, the
main entity to exclude is a cerebellopontine cis-
tern mass lesion. In the absence of additional
intratympanic or retrotympanic mass and the sus-
picion of arterial [reduced by pressure on the
ipsilateral internal carotid artery (ICA)] or venous
(reduced by pressure on the ipsilateral jugular vein)
aetiologies are particularly important.related to drugs, systemic processes (e.g. hyper-
PT. The choice and focus of imaging for PT is guidedThe prevalence of persistent tinnitus (lasting
more than 5 min) in the UK adult population is
around 10%; half of these patients find the symp-
tom moderately or severely annoying.
1
PT is much
less common than non-PT, affecting approximately
4% of patients with tinnitus.
2
PT may not require
radiological investigation; cases may be transient,
tension, anaemia, pregnancy), or other conditions,
gliomas, dural arteriovenous fistulae (dAVFs), idio-
pathic intracranial hypertension (IIH), venous
anatomical variations, and atheromatous arterial
disease represent the most frequent causes.
4e9
An underlying cause is usually identified in patients
with objective tinnitus (Table 1).
4e9
This review focuses on the radiological imaging of
by the clinical findings. The presence of a visiblea
Radiology Department, St Mary’s Hospital, London
King’s College Hospital, London, UK
Received 11 January 2008; received in revised form 17 J
Tinnitus may be continuous or pulsatile. Vascular lesions
pulsatile tinnitus. These include congenital vascular anom
and a variety of acquired vasculopathies. The choice of im
otoscopy, thin-section computed tomography (CT) is indic
possible imaging approaches. However, combined CT ang
ª 2008 The Royal College of Radiologists. Published by Els
Introduction
Tinnitus is the perception of an auditory sensation,
most frequently a ringing sound, in the absence of an
external stimulus. Tinnitus may be classified as
pulsatile (PT) or continuous. PT is usually related to
vascular causes and is pulse-synchronous (coinciding
with the patient’s heartbeat). It may be subjective
(heard only by the patient) or objective (also audible
to the examiner).K, and
b
Neuroradiology Department,
2008; accepted 1 August 2008
the most frequent radiologically demonstrable cause of
ies (which may be arterial or venous), vascular tumours,
ng depends on the clinical findings. If a mass is present at
d. In the otoscopically normal patient, there is a range of
raphy and venography is particularly useful.
r Ltd. All rights reserved.
radiological investigation of non-PT is low and
thin-section, T2-weighted magnetic resonance im-
aging (MRI) sequences are generally used for
screening.
3
Thus knowledge of the nature of the
tinnitus (PT versus non-PT) is essential.
There is a wide variation in the reported in-
cidence of structural abnormalities in patients
with PT ranging from 44e91% (Table 1).
4e9
This is
likely to reflect variations in the study populations,
expertise, and methods of investigation. Paragan-s

atie
ez
5%
320 G. Madani, S.E.J. ConnorTable 1 Reported incidence of structural abnormalities in p
Author Waldvogel
4
Sonm
Total no. of patients (Percentage
with objective tinnitus)
84 (42%) 74 (1
Investigations
aVarious imaging strategies have been proposed for
the investigation of PT in the otoscopically normal
patient and they continue to evolve. MRI (with
gadolinium), MR angiography (MRA), MR venogra-
phy (MRV), carotid ultrasound, CT with and without
contrast medium, and conventional angiography
have all been used rather inconsistently in
Ultrasound 68 12
Computed tomography 26 72
Magnetic resonance imaging 33 7
Magnetic resonance angiography 7 7
Selective angiography 46 5
CTA/V
Cause found 57 (68%) 50 (68%)
Vascular anomaly
Aberrant ICA 1 (1%)
Dehiscent jugular bulb 3 (4%)
High-riding jugular bulb 21 (28%)
JB/transverse sinus diverticulum 1 (1%)
Enlarged cortical draining vein
Vascular loop 1 (1%)
Vascular tortuosity
Dominant venous system
b
Acquired vasculopathy
Dural AVF 17 (20%) 2 (3%)
Pial AVF
Carotico-cavernous fistula 6 (7%)
Atheromatous ICA disease 7 (8%) 16 (22%
Fibromuscular dysplasia 5 (6%)
ICA aneurysm 1 (1%) 3 (4%)
ICA dissection
Extracranial AVF/M
Venous sinus thrombosis 1 (1%)
Tumour
Paraganglioma 5 (6%) 2 (3%)
Meningioma 1 (1%)
Other 1 (1%) 1 (1%)
Idiopathic intracranial hypertension 4 (5%)
Venous sinus stenosis 1 (1%)
Other
Otospongiosis
Myoclonus
Systemic causes 1 (1%)
No aetiology found in patients
with objective tinnitus (%)
7 (8%) 0
CTA/V, computed tomography angiography/venography; ICA, inte
a
Some patients underwent multiple investigations.
b
Association with the venous sinus dominance is speculative.
c
Four other patients had radiographic features of idiopathic intnts investigated for all causes of pulsatile tinnitus
5
Remley
6
Krishnan
7
Dietz
8
Sismanis
9
) 100 (25%) 16 (6%) 49 (33%) 145 (8%)previous patient series. Combined CT angiography
and venography (CTA/V) may be performed with
100 ml contrast medium injected at 3e4 ml/s and
a fixed delay of 25 s using contemporary multisec-
tion CT. This approach shows considerable promise
and has the advantage of demonstrating arterial,
venous, skull-base, and middle-ear disease entities
Not stated
69 10
24 49
49
68 17
16
80 (80%) 7 (44%) 28 (57%) 132 (91%)
8 (8%)
5 (5%) 1 (2%)
7 (7%) 1 (6%)
1 (1%) 1 (6%) 2 (4%)
1 (1%) 1 (2%)
1 (1%)
6 (4%)
6 (38%)
15 (15%) 3(%)
9 (18%)
1 (2%)
) 5 (5%)
4 (4%) 2 (4%)
2 (2%) 2 (1%)
1 (1%)
1 (2%) 1 (1%)
1 (2%)
25(25% 17 (12%)
2 (2%) 5 (10%)
2 (2%) 1 (1%)
2 (4%) 56
c
(39%)
1 (6%)
4 (3%)
1 (1%)
10 (8%)
0 Not stated 0 Not stated
rnal carotid artery; AVF, arteriovenous fistulae.
racranial hypertension but declined lumbar puncture.