Tinnitus Atlas
Tinnitus Atlas · Investigations and Audiological Evaluation in Tinnitus · Module 13

13CT and Vascular Imaging for Pulsatile Tinnitus

Pulsatile tinnitus is the one tinnitus that frequently has a findable, fixable cause — this module walks the imaging staircase from temporal-bone CT through CTA/CTV and MRA/MRV to catheter angiography, and shows how the sound's character points the radiologist to the right pathology.

FWhy pulsatile tinnitus earns its own imaging pathway

Most tinnitus is subjective, non-pulsatile, and never needs a scan. Pulsatile tinnitus is different: a rhythmic sound beating in time with the heart usually means real, turbulent or transmitted blood flow somewhere near the ear, and a meaningful minority of those patients harbour a structural lesion that is treatable — or, occasionally, dangerous if missed [2003].

The first fork is whether the noise is objective (the examiner can hear it with a stethoscope over the mastoid, neck or orbit) or subjective (only the patient hears it). The second fork is whether it is arterial (continuous, machinery-like, may have a bruit) or venous (a softer hum that quietens with ipsilateral neck compression or contralateral head turn). These bedside features steer which scan to order first and what the radiologist hunts for [2024].

THigh-resolution temporal-bone CT: the bone detectives

Thin-section (≤0.6 mm) temporal-bone CT is the workhorse for the objective, often conductive-flavoured pulsatile patient because it exquisitely shows bone. It detects superior semicircular canal dehiscence (a thin or absent bony roof producing a third-window pulse and autophony), a high-riding or dehiscent jugular bulb projecting into the middle ear, an aberrant internal carotid artery coursing through the tympanic cavity, a persistent stapedial artery, and the moth-eaten erosion of a glomus tympanicum or jugulare paraganglioma [2024].

CT also reveals the sigmoid sinus wall anomalies — dehiscence and diverticulum — that have emerged as a leading cause of venous pulsatile tinnitus, and the bony signs of idiopathic intracranial hypertension such as an empty sella and dilated optic nerve sheaths. Reading temporal-bone CT for pulsatile tinnitus is therefore a structured checklist, not a glance.

The pulsatile-tinnitus imaging staircase

increasing invasiveness & costTemporal-bone CT (≤0.6 mm)bone: SSCD, jugular bulb, aberrant ICA, glomus erosion, sinus wallCTA + CTVlumen + bone: aberrant carotid, sinus diverticulum, transverse-sinus stenosisMRI brain/IAC + MRA/MRVsoft tissue + flow: dural-AVF clues, IIH signs, no radiationCatheter DSA — reference standarddynamic flow, fistula grading, treat in same sittingmost patientsstop here

Climb only as far as the clinical picture demands; most patients are answered by CT ± CTA/CTV. Schematic.

TCTA/CTV and MRA/MRV: imaging the flow

When the question shifts from bone to lumen, cross-sectional angiography takes over. CT angiography and venography (CTA/CTV) give superb arterial and venous luminal detail with bone in the same dataset, ideal for confirming an aberrant carotid, sigmoid sinus diverticulum, or transverse-sinus stenosis. Contrast-enhanced MR angiography and venography (MRA/MRV), paired with routine brain and internal-auditory-canal MRI, avoid ionising radiation and add soft-tissue and flow-related information — useful when a dural arteriovenous fistula or venous stenosis from intracranial hypertension is suspected [2008].

A dural arteriovenous fistula is the lesion no one wants to miss: it can present with isolated pulsatile tinnitus yet carry a risk of haemorrhage, so suggestive arterial-side findings — engorged transosseous feeders, early venous filling, asymmetric flow voids — mandate escalation rather than reassurance.

Arterial vs venous pulsatile tinnitus: a two-branch decision

Pulsatile tinnitus + auscultation/otoscopyARTERIAL cluesVENOUS cluesRetrotympanic mass / third-window → CT first
CluesContinuous machinery sound · Bruit louder · Neurological features · Does NOT change with neck compressionFirst scanCTA / MRA, low threshold for DSATarget lesionsDural AVF, glomus, aberrant ICA, persistent stapedial artery, carotid-cavernous fistula

The sound’s character chooses the first scan. Tap a branch. Schematic.

CCatheter DSA: the reference standard

Digital subtraction angiography (DSA) remains the gold standard for vascular pulsatile tinnitus because its dynamic, selective injections show the timing and direction of flow that static CT/MR snapshots cannot. It is the only reliable way to fully characterise a dural arteriovenous fistula (Cognard/Borden grading), to confirm or exclude a small fistula when non-invasive imaging is equivocal, and to plan or deliver endovascular treatment in the same sitting [2024].

Because DSA is invasive and carries a small stroke risk, it is reserved for cases where the bedside picture and non-invasive imaging point to an arterial/fistulous cause, or where strongly suspected pathology has not been explained. In idiopathic intracranial hypertension with transverse-sinus stenosis, DSA with manometry both confirms the pressure gradient and serves as the platform for venous stenting, which can abolish the tinnitus — although the noise may persist or recur in some patients [2025][2025].

Temporal-bone CT findings in pulsatile tinnitus

SSCDJugular bulbAberrant ICAGlomusSigmoid sinuscoronal temporal bone (right ear, schematic)
Tap a hotspot for the finding to look for.

One thin-section CT, five questions to ask. Red = arterial/other; blue = venous. Schematic.

CPutting it together: an arterial-versus-venous algorithm

A practical pathway begins with otoscopy and auscultation. A retrotympanic mass or objective bruit, conductive picture, or third-window symptoms favour temporal-bone CT first. A soft venous hum that compresses away in a patient with headache and visual symptoms favours MR/MRV with attention to the sinuses and signs of raised pressure. A continuous arterial bruit, especially with any neurological feature, pushes toward CTA/MRA and a low threshold for DSA to exclude a fistula [2008].

Even a thorough work-up leaves a proportion of pulsatile tinnitus unexplained, and a focused, character-driven approach maximises yield while sparing low-risk patients unnecessary radiation and contrast [2025].

Case 5.13
A 41-year-old woman with a BMI of 34 reports a whooshing noise in her right ear, in time with her pulse, for three months. It is louder when she lies down, softer when she presses on the right side of her neck, and she has had morning headaches and brief greying-out of vision on standing. Otoscopy is normal; no retrotympanic mass. Auscultation reveals a soft hum over the right mastoid that disappears with light jugular compression.

What is the most appropriate first-line imaging strategy?

Self-assessment — Module 133 questions
Question 1 · Foundation

Which bedside manoeuvre classically reduces a venous pulsatile tinnitus?

Question 2 · Trainee

Why is catheter DSA still considered the reference standard for vascular pulsatile tinnitus?

Question 3 · Clinician

A continuous machinery-like bruit with early venous filling on imaging in a pulsatile-tinnitus patient should raise suspicion for which lesion?

Tracked locally in your browser — see /progress for the dashboard.