MRI Middle Ear and Petrous Bone — Generic Standard Protocol

MRIninja Knowledge Base | Master / General Protocol Page Related pages: MRI CPA and Inner Ear Generic Standard Protocol Version 1.0 — May 2026

1. Executive Summary

MRI of the middle ear and petrous bone occupies a distinct and complementary role to high-resolution CT (HRCT) of the temporal bone. HRCT remains the gold standard for bony anatomy — tympanic membrane, ossicular chain, bony labyrinth, air cell system, tegmen integrity, and sigmoid sinus plate — and is the primary investigation for most middle ear disease. MRI, however, provides unique soft tissue characterisation that HRCT cannot deliver: the distinction between cholesteatoma and non-cholesteatomatous middle ear soft tissue; the characterisation of petrous apex lesions; the assessment of the facial nerve along its entire intratemporal course; and the evaluation of vascular anomalies and glomus tumours. The two modalities are therefore complementary rather than competitive.

The middle ear and petrous bone MRI protocol shares some sequences with the CPA/inner ear protocol (see companion page), but has distinct primary targets: the epitympanum and mesotympanum (cholesteatoma); the petrous apex (various cystic and solid lesions); the intratemporal facial nerve (CN VII segments); and the jugular foramen region (glomus jugulare, glomus tympanicum). The defining sequence for middle ear MRI is non-EPI DWI (EPI-DWI is insufficient for the 2–4 mm middle ear structures), which requires dedicated non-echo-planar diffusion-weighted sequences (PROPELLER/BLADE DWI, HASTE DWI, or MR cholesteatoma protocols using dedicated non-EPI acquisitions) to detect cholesteatoma with sufficient spatial resolution and without the geometric distortion of standard EPI.

1.1 Core Strengths

Cholesteatoma detection and characterisation: the most important MRI contribution to middle ear assessment. Cholesteatoma — an epidermoid cyst of the middle ear space — contains keratinous debris that restricts diffusion (ADC 0.4–0.7 × 10⁻³ mm²/s). Standard EPI-DWI has insufficient resolution and too much distortion for reliable cholesteatoma detection (the target lesion may be 2–5 mm). Non-EPI DWI (PROPELLER/BLADE, HASTE, or equivalent) provides the spatial resolution (1–2 mm) and freedom from geometric distortion required for reliable detection. MRI is the primary modality for cholesteatoma characterisation before primary surgery and for recurrence/residual detection after canal wall up or canal wall down tympanoplasty.

Petrous apex lesion characterisation: the petrous apex contains marrow fat, air cells, and is contiguous with the petrous bone cortex and the clivus. Multiple distinct pathologies affect this region and are reliably characterised by MRI: - Cholesterol granuloma: T1-bright (blood degradation products), T2-bright, non-enhancing — the most common petrous apex lesion - Mucocele / trapped fluid: T1-variable, T2-bright, non-enhancing - Cholesteatoma of petrous apex: T1-dark, T2-bright, DWI restricts - Petrous apex effusion (asymmetric marrow replacement): T2-bright, T1-variable, no enhancement

CT identifies these lesions but cannot reliably distinguish their soft tissue nature.

Facial nerve (intratemporal) assessment: the intratemporal segments of CN VII — labyrinthine segment (7 mm), geniculate ganglion, tympanic (horizontal) segment (12 mm), mastoid (vertical) segment (15 mm), and stylomastoid exit — are fully characterised by post-contrast T1. Enhancement of the geniculate ganglion (present in 10–15% of normals), abnormal enhancement of the tympanic or mastoid segments, and facial nerve masses (schwannoma, hemangioma) are detected. This assessment extends the CPA/inner ear protocol to the full intratemporal facial nerve course.

Jugular foramen and vascular lesions: glomus jugulare (paraganglioma of the jugular foramen) and glomus tympanicum (paraganglioma of the middle ear) show the “salt-and-pepper” T2 appearance from tumour vascularity and internal flow voids, with intense enhancement. MRI characterises these lesions definitively and assesses their relationship to the jugular bulb, internal carotid artery, and facial nerve. CT delineates bony destruction; MRI provides vascular and soft tissue characterisation.

Post-operative middle ear assessment: after tympanoplasty, mastoidectomy, or cochlear implantation, MRI assesses residual or recurrent cholesteatoma (DWI), graft integrity, and any complication (meningitis, labyrinthine fistula).

1.2 Intrinsic Limitations of the Generic Protocol

Ossicular chain assessment: MRI cannot assess the ossicular chain (malleus, incus, stapes) with the reliability of HRCT. The tiny ossicles (< 3 mm) and their precise relationships (incudostapedial joint, footplate mobility) require CT. MRI provides soft tissue context but not ossicular detail.

Bony labyrinthine anatomy: the bony cochlear architecture, scala tympani and vestibuli dimensions, semicircular canal dimensions, and modiolus detail for surgical planning (CI candidacy) require HRCT. MRI provides soft tissue and fluid characterisation within these structures but not bony architecture.

Small cholesteatoma at the resolution limit: non-EPI DWI at 1–2 mm resolution detects cholesteatoma ≥ 3 mm reliably. A 1–2 mm cholesteatoma residual in an operated middle ear may be below the detection threshold. The clinically relevant threshold (requiring re-operation) is generally 3–5 mm, which is within the detection capability of modern non-EPI DWI.

Tympanic membrane and ossicular mobility: MRI provides no functional assessment of these structures.

When dedicated child protocols are required: cholesteatoma primary detection and staging; cholesteatoma post-operative recurrence/residual detection; facial nerve palsy with intratemporal assessment; glomus tumour (paraganglioma) characterisation and staging; petrous apex lesion characterisation; Gradenigo syndrome (petrous apex osteitis); tegmen tympani defect and CSF leak assessment; cochlear implant complications.

2. Main Clinical Indications

2.1 Standard Indications

Cholesteatoma evaluation is the primary indication unique to this protocol. When HRCT demonstrates soft tissue in the middle ear with bony erosion, MRI is obtained to: (a) characterise whether the soft tissue represents cholesteatoma (DWI-positive) or inflammatory middle ear disease (granulation tissue, polyp, mucoid effusion — DWI negative); (b) assess the extent of cholesteatoma into the labyrinth, tegmen, or posterior fossa; (c) evaluate intracranial complications. The generic protocol with non-EPI DWI is appropriate for initial cholesteatoma assessment. Post-operative follow-up requires a dedicated protocol emphasising minimal echo-spacing non-EPI DWI for maximal sensitivity.

Facial nerve palsy (intratemporal course) — when the intratemporal segment of CN VII is the clinical concern (recurrent Bell’s palsy, suspected facial nerve schwannoma, traumatic facial palsy, facial nerve schwannoma known or suspected), post-contrast T1 through the full intratemporal course is the primary sequence. The generic protocol covers this adequately. For surgical planning of facial nerve schwannoma, dedicated thin-slice post-contrast T1 along the entire facial canal trajectory is required.

Petrous apex lesion — any petrous apex abnormality detected on HRCT or on another brain MRI. The T1 (pre-contrast) is the pivotal sequence for characterising cholesterol granuloma (T1-bright). The full MRI protocol (T1 pre-contrast + T2 + post-contrast T1) characterises all petrous apex entities. The generic protocol is appropriate for initial characterisation.

Pulsatile tinnitus with glomus tumour suspicion — glomus tympanicum appears as a middle ear vascular mass (pulsatile tinnitus, visible behind the tympanic membrane on otoscopy). MRI characterises the lesion (T2 “salt-and-pepper” + intense enhancement) and distinguishes it from a high jugular bulb, aberrant internal carotid artery, or dehiscent jugular bulb — all of which produce pulsatile tinnitus and a middle ear vascular appearance. The generic protocol with contrast is appropriate; if paraganglioma staging is required, supplementary neck MRI is indicated.

Jugular foramen mass — glomus jugulare, schwannoma (CN IX, X, XI), meningioma, and neurofibroma all arise in or extend to the jugular foramen. MRI characterises the mass and defines its relationship to the jugular vein, sigmoid sinus, internal carotid artery, and the lower cranial nerves. The generic protocol is appropriate; for extensive paraganglioma staging, a whole-neck survey may be required.

Post-operative middle ear/mastoid assessment — assessment for residual/recurrent cholesteatoma after tympanoplasty (canal wall up or canal wall down). This is the most technically demanding indication because the residual cholesteatoma may be very small (2–5 mm) in a post-operative middle ear filled with inflammatory tissue, fat graft, and prosthetics. The generic protocol with non-EPI DWI is the appropriate starting point; a dedicated post-operative cholesteatoma protocol with optimised non-EPI DWI is the standard for this indication.

2.2 Urgent Red Flags Requiring Expedited or Emergency Imaging

3. Preparation Reference

Universal MRI safety screening belongs to the general MRI preparation page and is not repeated here.

3.1 Anatomy-Specific Preparation Items

The preparation requirements for middle ear and petrous bone MRI are nearly identical to the CPA/inner ear protocol (see companion page). Key items specific to the middle ear and petrous bone context:

Cochlear implants: as documented in the CPA/inner ear page — mandatory model verification and MRI compatibility confirmation before any temporal bone MRI. The closer proximity of the imaging target (middle ear) to the implant (typically in the mastoid process) means susceptibility artefact from the implant may significantly affect the mastoid portion of the examination. Document and note in the report.

Middle ear prostheses (titanium TORP/PORP): as per CPA/inner ear page. Titanium prostheses are MR-compatible. However, their susceptibility artefact in the epitympanum may degrade the non-EPI DWI precisely in the region of primary cholesteatoma concern. This is an inherent limitation — document in the report.

Post-operative history is critical: the entire interpretation of this examination depends on knowing the surgical history. Specifically: - Canal wall up (CWU) vs canal wall down (CWD) tympanoplasty: determines the expected anatomy and where residual/recurrent cholesteatoma would be located - Fat graft placement: fat grafts in the middle ear or mastoid are T1-bright and STIR-suppressed — must not be misidentified as cholesterol granuloma - Obliterative mastoidectomy with bone chips vs fat obliteration: significantly changes the T1 appearance - Prior DWI MRI result: prior negative DWI does not exclude new recurrence; prior positive DWI with interval size change monitors treatment

3.2 Patient Positioning on the MRI System

Identical to the CPA/inner ear protocol: supine, head-first, head coil, isocentre at the external auditory canal level (tragus level). See the companion CPA/inner ear page for full detail.

The key positioning distinction for middle ear vs inner ear assessment: the target anatomy (middle ear cleft, epitympanum, mastoid) is slightly more lateral and superficial than the IAC. The B0 shim volume should be centred on the temporal bones bilaterally. Head coil isocentre at the tragus level is identical for both protocols and covers both regions.

4. Standard Protocol Design

4.1 Mandatory Core Sequences

4.2 Conditional Sequences

4.3 Rationale Summary Per Sequence

T1-weighted TSE (non-fat-suppressed) — the pivotal sequence for petrous apex pathology. The pre-contrast T1 must be acquired without fat suppression because: - Cholesterol granuloma is T1-bright (methaemoglobin, blood degradation products) — its T1 hyperintensity is pathognomonic and is the primary characteristic distinguishing it from other petrous apex lesions - Fat grafts in obliterative mastoidectomy are T1-bright — knowing the surgical history and identifying fat grafts on T1 prevents misinterpretation - Marrow fat of the petrous apex is normally T1-bright — asymmetric marrow signal (T1-dark = marrow replacement, T1-bright = fat) is evaluated against the normal side

T2-weighted TSE provides: - Soft tissue characterisation of middle ear content (cholesteatoma: T2-bright homogeneous; granulation tissue: T2-bright heterogeneous; effusion: T2-very bright) - Petrous apex T2 signal (cholesterol granuloma: T2-bright; cholesteatoma: T2-bright; effusion: T2-very bright) - Posterior fossa survey (meningitis, lateral sinus thrombosis — low signal in thrombosed sinus) - Brain parenchyma adjacent to the temporal bone

Non-EPI DWI (PROPELLER/BLADE or equivalent) — the most protocol-specific sequence. Standard EPI-DWI as used in brain DWI protocols is inadequate for middle ear cholesteatoma because: 1. EPI geometric distortion at 3T: the air-bone interfaces of the middle ear (tympanic membrane, ossicular chain, mastoid air cells) produce massive B0 disturbance → EPI geometric distortion of 5–15 mm → the DWI “lesion” may be displaced 5–15 mm from the true lesion location → false localisation 2. EPI resolution: standard brain DWI (5 mm × 5 mm voxels) is insufficient for a 2–5 mm cholesteatoma

Non-EPI DWI alternatives: - PROPELLER/BLADE DWI (GE/Siemens): multi-shot rotating k-space acquisition; inherently motion-corrected; no EPI geometric distortion; resolution 1.5–2 mm in-plane - HASTE DWI (single-shot TSE with diffusion): relatively fast; moderate resolution - Readout-segmented EPI (rs-EPI, RESOLVE): significantly reduced distortion compared with single-shot EPI; achievable resolution 1.5–2 mm; increasingly used as a compromise between standard EPI (too distorted) and PROPELLER (slower)

The choice of non-EPI DWI technique is vendor- and institution-dependent. The critical requirement is: in-plane resolution ≤ 2 mm; geometric distortion at the epitympanum/mastoid < 2 mm; ADC map generation.

ADC in cholesteatoma: approximately 0.4–0.8 × 10⁻³ mm²/s (restricted; comparable to epidermoid cyst). Post-operative granulation tissue and mucoid effusion do not restrict diffusion (ADC > 1.5 × 10⁻³ mm²/s). This ADC difference is the basis for non-invasive cholesteatoma detection.

Post-contrast T1 fat-suppressed (axial + coronal) provides: - Intratemporal facial nerve enhancement: labyrinthine segment, geniculate ganglion, tympanic segment, mastoid segment - Middle ear and mastoid enhancement: cholesteatoma does not enhance (critical negative finding); granulation tissue enhances uniformly; glomus tumour enhances intensely - Glomus tympanicum/jugulare: intense enhancement with internal flow voids (“salt-and-pepper” on T2) - Cholesterol granuloma: may enhance peripherally; central T1-bright content does not change significantly with gadolinium - Perineural spread: tracking enhancement along CN VII or lower cranial nerves

The coronal post-contrast T1 is specifically valuable for: (a) the geniculate ganglion (normally may enhance mildly in 10–15% of the population; abnormal if associated with facial palsy); (b) the tympanic and mastoid segments of CN VII; (c) the tegmen defect assessment.

3D CISS (0.5–0.7 mm isotropic): shared with the CPA/inner ear protocol. For the middle ear and petrous bone context, the 3D CISS is specifically used for: - Cholesteatoma extension into the labyrinth (intralabyrinthine cholesteatoma — T2-dark filling defect in the normally T2-bright cochlea/semicircular canals) - Relationship of a petrous apex lesion to the labyrinth and IAC - CSF otorrhoea: at steady state, CSF in the middle ear may be identified as T2-bright on CISS (post-gadolinium CISS or dedicated cisternography)

4.4 Sequence Matching and Cross-Sequence Consistency

Pre-contrast T1 and post-contrast T1 must use identical geometry for accurate lesion characterisation. The T1-bright content of a cholesterol granuloma should appear identical pre- and post-contrast (does not enhance meaningfully); a solid mass (meningioma, schwannoma) will show clear post-contrast signal increase. This comparison is only possible with matched geometry.

The non-EPI DWI must be registered (mentally or by software) with the T1 and T2 images to confirm that a DWI-positive focus corresponds to the anatomical region of concern identified on T1/T2. For post-operative assessment, the DWI finding must be co-localised with the surgical anatomy on T1/T2 before clinical reporting.

4.5 Fat Suppression

T1 pre-contrast is acquired without fat suppression — the T1-bright signal from fat (normal marrow, fat graft), cholesterol granuloma, and blood products is diagnostically essential and must not be suppressed.

Post-contrast T1 uses fat suppression (SPAIR or Dixon) — the post-contrast T1 is acquired with fat suppression to: (a) suppress T1-bright marrow fat and fat grafts that would otherwise obscure enhancing tissue; (b) improve the contrast-to-noise ratio for small enhancing lesions at the facial nerve. At isocentre with a head coil, SPAIR provides reliable fat suppression. Dixon is the preferred alternative at 3T.

STIR post-gadolinium is contraindicated as throughout all MRIninja protocols.

5. Optimisation Strategy

5.1 Artifact Reduction by Source

Air-tissue susceptibility at the middle ear (primary artefact source): the middle ear is an air-filled cavity surrounded by bone and soft tissue. The air-bone-soft tissue interfaces produce the most severe B0 disturbances in the head. These cause: - EPI geometric distortion: as described — this is the reason non-EPI DWI is mandatory - Fat suppression failure adjacent to the middle ear: SPAIR may fail at the epitympanum and around the mastoid air cells. Dixon is more robust. For sequences adjacent to the middle ear air spaces, non-uniform fat suppression is expected and should be noted in the report. - T2* signal loss at the tympanic membrane region on GRE sequences

Pulsatile flow artefact from jugular bulb and sigmoid sinus: the jugular bulb and sigmoid sinus produce pulsatile flow artefacts in the phase-encoding direction. These appear as periodic bright bands. A-P phase encoding displaces them anteroposteriorly. Saturation bands over the sigmoid sinus and jugular vein can reduce this artefact for temporal bone sequences.

Post-surgical metalwork: titanium prostheses, PORP/TORP, bone-anchored hearing aid posts, and any remaining surgical hardware produce susceptibility artefacts that directly affect the middle ear region. The size of the artefact at 1.5T is considerably smaller than at 3T. For complex post-operative cases with extensive metalwork, 1.5T may provide a better non-EPI DWI result than 3T.

Cholesteatoma matrix enhancement (pitfall, not artefact): cholesteatoma can show peripheral enhancement of the sac wall on post-contrast T1. This is NOT a sign that the lesion is “not a cholesteatoma” — it reflects enhancement of the surrounding fibrous tissue. The critical finding is the non-enhancing central keratin core on DWI. This is an interpretive pitfall, not a technical artefact.

5.2 Protocol Efficiency and Throughput

A complete middle ear and petrous bone MRI — T1 axial + T2 axial + non-EPI DWI + post-contrast T1 (axial + coronal) + 3D CISS — requires approximately 30–40 minutes at 3T.

For cholesteatoma-specific protocols (primary or post-operative), the non-EPI DWI is the most time-consuming component because PROPELLER/BLADE acquisitions are inherently slower than EPI (requiring multiple rotating shots). At 3T: PROPELLER DWI ≈ 4–6 minutes for adequate brain coverage at 2 mm resolution.

Rs-EPI (RESOLVE/MUSE) as a compromise provides 1.5–2 mm resolution with substantially reduced distortion in 2–3 minutes per acquisition — a clinically attractive balance when PROPELLER is not available.

5.3 Field Strength Considerations

3T is preferred for: sub-millimetre 3D CISS resolution; superior non-EPI DWI resolution; higher SNR for fine middle ear soft tissue characterisation.

1.5T may be preferred for: post-operative patients with extensive metalwork (titanium prostheses, cochlear implant) where susceptibility artefact at 3T is diagnostically limiting; patients where EPI geometric distortion management is critical and PROPELLER is not available.

The key 3T limitation for middle ear MRI: the air-bone interfaces of the middle ear produce larger susceptibility artefacts at 3T than at 1.5T. Non-EPI DWI at 3T still shows more distortion near the middle ear than at 1.5T (though substantially less than EPI). For very small post-operative cholesteatoma (< 3 mm) in a distorted post-operative cavity, 1.5T PROPELLER DWI may provide better lesion localisation than 3T RESOLVE.

6. Contrast Use Principles Specific to Middle Ear and Petrous Bone MRI

6.1 Non-Contrast Standard Protocol — Sufficient For

Non-contrast middle ear MRI (T1 + T2 + non-EPI DWI + 3D CISS) is diagnostically adequate for: - Primary cholesteatoma characterisation (DWI is the primary sequence; contrast adds little for typical cases where DWI is clearly positive) - Petrous apex cholesterol granuloma (T1-bright on pre-contrast T1 is pathognomonic; contrast not required for diagnosis) - Congenital cholesteatoma (non-EPI DWI positive; no prior surgery; diagnosis based on DWI)

6.2 Gadolinium Indicated — Region-Specific Contexts

Post-contrast T1 is required for: - Facial nerve assessment (Bell’s palsy, facial nerve schwannoma, perineural spread): post-contrast T1 through the full intratemporal facial canal is the primary sequence for all facial nerve indications - Glomus tumour (paraganglioma): intense enhancement + flow voids (“salt-and-pepper”) is the characteristic appearance requiring contrast - Post-operative cholesteatoma assessment: granulation tissue vs recurrent cholesteatoma distinction — granulation tissue enhances; cholesteatoma does not enhance - Intracranial complication of mastoiditis: meningeal enhancement; lateral sinus thrombosis (absent enhancement in thrombosed sinus) - Cholesteatoma with uncertain DWI: if non-EPI DWI is equivocal, post-contrast T1 provides additional information (absence of central enhancement supports cholesteatoma) - Petrous apex mass characterisation (other than cholesterol granuloma): enhancement pattern helps distinguish mucocele (non-enhancing) from schwannoma (enhancing) vs chordoma (heterogeneous enhancement)

6.3 Post-Contrast Acquisition Timing

Standard equilibrium phase (3–5 minutes post-injection) for all middle ear and petrous bone post-contrast sequences. No specific arterial or delayed phase is required for the standard protocol.

7. Reporting Essentials

7.1 Interpretation Framework

Middle ear and petrous bone MRI reporting follows a systematic compartmental approach:

Middle ear space (bilateral): - Air vs soft tissue content: normal air / soft tissue (opacification) - Soft tissue characterisation: DWI-positive (cholesteatoma); DWI-negative (granulation tissue, effusion, granuloma) - Enhancement of middle ear soft tissue: enhancing (granulation, glomus) / non-enhancing (cholesteatoma, effusion)

Facial nerve canal (bilateral): - Labyrinthine, geniculate ganglion, tympanic, mastoid segments: normal / abnormal signal / enhancement - Normal geniculate ganglion enhancement (10–15% of normals) — bilateral mild enhancement without facial palsy: note as normal variant - Abnormal: asymmetric or extensive enhancement in the context of facial palsy; mass enlargement of any segment

Petrous apex (bilateral): - T1 signal: normal fat / T1-bright (cholesterol granuloma, haemorrhage) / T1-dark (marrow replacement, cholesteatoma) - T2 signal - Enhancement - Relationship to carotid canal and IAC

Mastoid air cells and mastoid cortex: - Pneumatisation: well-pneumatised / hypopneumatised / sclerotic - Air cell signal: air / fluid-filled / soft tissue - Cortical integrity: intact / erosion

Tegmen tympani and posterior cranial fossa plate: - Integrity: intact / defect (CT better for bony detail; MRI for associated brain herniation or CSF)

Jugular foramen (bilateral): - Mass: glomus jugulare; schwannoma; meningioma

Sigmoid sinus and jugular bulb: - Signal: normal flow void / thrombosis (T2-bright; no enhancement in the thrombosed segment)

7.2 Mandatory Reporting Checklist

Technical quality: - [ ] Field strength; coil; DWI technique used (EPI vs non-EPI); specify vendor technique name - [ ] Motion: non-EPI DWI quality adequate for 2–5 mm lesion detection - [ ] Fat suppression: uniform / failure at middle ear region

Right side: - [ ] Middle ear soft tissue: absent (normal) / present - If present: DWI — restricted / not restricted; ADC value if measured; enhancement — yes/no - [ ] Facial nerve canal: all segments assessed / limited (artefact); enhancement: normal / abnormal - [ ] Petrous apex: normal / abnormal — specify T1/T2/enhancement characteristics - [ ] Mastoid: air-filled / partially / completely opacified; cortex intact

Left side: same checklist

Bilateral structures: - [ ] Jugular foramina: normal / mass - [ ] Sigmoid sinuses: normal / thrombosis signal - [ ] Posterior cranial fossa: normal / abnormal

7.3 Structured Reporting

Reports must include: Indication (cholesteatoma primary/post-operative / facial palsy / petrous apex / pulsatile tinnitus / other); Technique (field strength, coil, sequences, DWI type: EPI/non-EPI/PROPELLER/RESOLVE, contrast if used); Surgical history (if relevant — type of tympanoplasty, date, side); Comparison (prior MRI date); Findings (bilateral systematic as per checklist); DWI result (restricted / not restricted; location; size; ADC); Impression (primary finding; DWI conclusion regarding cholesteatoma); Limitations (metalwork artefact; EPI distortion if standard DWI used; fat suppression failure).

7.4 Incidental Findings — Clinical Decision Framework

Usually benign: asymmetric mastoid pneumatisation; small amount of fluid in mastoid air cells on the dependent side (positional, not inflammatory); high jugular bulb (vascular variant — produces pulsatile tinnitus; important to recognise to avoid surgical inadvertent trauma).

May require follow-up: petrous apex lesion with intermediate T1 signal (not clearly diagnostic for cholesterol granuloma) — correlation with CT and short-interval follow-up or ENT evaluation; incidental enhancement of the geniculate ganglion in a patient without facial palsy — note as possible normal variant; monitor clinically.

Urgent communication: cholesteatoma with intracranial extension (epidural abscess, meningitis signs); lateral sinus thrombosis; sigmoid sinus thrombophlebitis with septic emboli; unexpected malignant lesion involving the temporal bone or jugular foramen.

8. MRI Technologist Pearls

8.1 Sequence Order Logic

1. Three-plane localiser
2. T1 axial ← pre-contrast; petrous apex characterisation first
3. T2 axial
4. 3D CISS ← before contrast; labyrinthine assessment; inner ear component
5. Non-EPI DWI ← cholesteatoma primary sequence; before contrast
6. Contrast injection (if indicated)
7. Post-contrast T1-FS axial
8. Post-contrast T1-FS coronal ← facial nerve segments

Non-EPI DWI is placed before contrast injection because: (a) gadolinium may (theoretically) shorten T1 of the cholesteatoma wall, potentially altering DWI signal indirectly; (b) this is standard practice to ensure a clean DWI acquisition.

8.2 Positioning Tricks

For patients with post-operative middle ear anatomy: review the operative note before positioning to understand which side was operated and what hardware was placed. Position the head slightly more toward the non-operated side if unilateral hardware is the primary artefact source — placing the hardware slightly off-axis reduces its impact on the target anatomy.

For non-EPI DWI (PROPELLER/BLADE): the patient must be completely still during this acquisition. The PROPELLER readout collects multiple rotating k-space blades — patient motion between blades creates incoherence that degrades the image. A specific verbal instruction (“This is the most important sequence — please stay completely still for the next 5 minutes”) before the DWI can meaningfully improve image quality.

8.3 Fast Salvage Protocol

Approximately 10–11 minutes — covers the three primary diagnostic questions for cholesteatoma detection and facial nerve assessment.

8.4 Common Avoidable Errors

9. Quality Control Checklist

• T1 axial pre-contrast: acquired WITHOUT fat suppression; petrous apex signal documented
• T2 axial: middle ear space and mastoid covered bilaterally
• 3D CISS: inner ear and labyrinthine structures visible bilaterally (CISS quality check per CPA/inner ear protocol)
• Non-EPI DWI type documented (PROPELLER/BLADE / RESOLVE / HASTE / other): adequate resolution for ≤ 3 mm lesion detection
• Non-EPI DWI: epitympanum and mastoid covered bilaterally (check on axial coverage)
• Non-EPI DWI: ADC map generated
• Post-contrast T1-FS: fat suppression quality at middle ear level — uniform / failure documented
• Post-contrast T1-FS coronal: geniculate ganglion level included bilaterally
• Post-contrast T1-FS: middle ear soft tissue enhancement status documented
• Bilateral temporal bones within FOV
• Post-surgical anatomy (if applicable): surgical cavity within DWI and T1 coverage

11. Evidence Gaps and Ongoing Debate

Non-EPI DWI technique standardisation (PROPELLER vs RESOLVE vs HASTE): multiple non-EPI DWI techniques are in clinical use for cholesteatoma detection, and no formal head-to-head comparison with surgical verification has established a single superior technique across all platforms. PROPELLER DWI is the most extensively validated (Radiology 2010 [1], Eur Radiol 2006 [3]); RESOLVE has demonstrated promise in limited series. The optimal non-EPI DWI approach remains platform-dependent and centre-dependent.

Minimum detectable cholesteatoma size: published sensitivities for non-EPI DWI use surgical series where the smallest detected lesion is typically 3–4 mm. The sensitivity for 1–2 mm residual cholesteatoma (the most clinically challenging post-operative scenario) has not been reliably established. Some expert centres report sensitivity approaching 90% for ≥ 3 mm residuals.

Role of delayed Gd-enhanced T1 for cholesteatoma: early protocols used delayed post-contrast T1 (at 45 minutes) to identify non-enhancing cholesteatoma surrounded by enhancing granulation tissue. This technique has been largely replaced by non-EPI DWI in expert centres, but some departments still use the delayed T1 technique in combination with DWI. Whether the combination of DWI + delayed Gd-T1 outperforms DWI alone for post-operative residual detection has not been established by a large comparative trial.

3T vs 1.5T for non-EPI DWI: while 3T provides higher SNR for thinner slices and higher resolution, the increased susceptibility at 3T produces larger distortion artefacts even with non-EPI techniques. For post-operative patients with extensive metalwork, 1.5T non-EPI DWI may provide better image quality than 3T. Formal comparative trials across different post-operative scenarios are lacking.