MRI Cerebellopontine Angle and Inner Ear — Generic Standard Protocol

MRIninja Knowledge Base | Master / General Protocol Page Version 1.0 — May 2026

1. Executive Summary

The cerebellopontine angle (CPA) and inner ear represent the most technically demanding routine MRI examination in clinical neuroradiology. The target structures — the cochlea, semicircular canals, vestibule, internal auditory canal (IAC), vestibulo-cochlear nerve, facial nerve, and their relationships to the brainstem and CPA cistern — are measured in millimetres (cochlear turns: 0.3–0.8 mm diameter) or sub-millimetres. No other routine clinical brain MRI demands sub-millimetre isotropic resolution as a standard requirement. The primary diagnostic sequences must simultaneously provide: (1) high fluid signal (CSF and perilymph) to outline the membranous labyrinth within the bony labyrinth; (2) high soft-tissue contrast to identify intracanalicular and CPA masses; (3) adequate T1 post-contrast sensitivity for enhancing lesions; and (4) sufficient DWI sensitivity for cholesteatoma.

MRI is the primary imaging modality for all suspected CPA and inner ear pathology in adults. High-resolution CT (HRCT) of the temporal bone is the complementary modality — CT is superior for bony labyrinth and ossicular detail, congenital bony anomalies, and fractures. For the IAC contents and soft tissue masses, MRI is the gold standard.

The defining technical characteristic of this protocol — which separates it from every other protocol in the MRIninja knowledge base — is the mandatory 3D heavily T2-weighted sequence (CISS/DRIVE/FIESTA-C) at sub-millimetre isotropic resolution (0.5–0.7 mm). This sequence provides the fluid signal necessary to visualise the membranous labyrinth structures and the nerves within the CSF of the IAC, and its quality determines whether the examination is diagnostic. The 3D CISS is not optional, not a conditional add-on, and not replaceable by any 2D sequence for the primary indications of this protocol.

1.1 Core Strengths

Sub-millimetre nerve and labyrinthine structure visualisation: 3D CISS/FIESTA-C at 0.5–0.7 mm isotropic resolves the individual fascicles of the vestibular and cochlear nerves within the IAC (the cochlear nerve: 0.8–1.2 mm diameter; superior and inferior vestibular nerves: 0.6–1.0 mm diameter). Nerve aplasia, hypoplasia, or compression by a small vestibular schwannoma are reliably detected at this resolution.

Membranous labyrinth assessment: the fluid-filled spaces of the cochlea, semicircular canals, and vestibule are directly visualised as T2-bright structures. Obliteration (fibrous or osseous labyrinthitis — labyrinthine ossification), congenital malformations, and labyrinthine fistulas are detected. CT detects only severe ossification; MRI detects early fibrous obliteration before bone forms.

CPA mass characterisation: vestibular schwannomas, meningiomas, epidermoid cysts, and other CPA masses are characterised by their signal characteristics on T2 (CISS), T1, and post-contrast T1. The relationship of the mass to the cochlear and facial nerves, its extension into the IAC, and any intralabyrinthine extension are defined by the 3D CISS.

Facial nerve mapping: the entire intratemporal course of CN VII — from the IAC (intracanalicular segment) to the stylomastoid foramen (mastoid segment) — is visualised on post-contrast T1 for enhancement (facial neuritis, perineural spread, schwannoma).

DWI for cholesteatoma: the DWI b=1000 sequence uniquely identifies cholesteatoma, which restricts diffusion (ADC approximately 0.4–0.7 × 10⁻³ mm²/s) and appears hyperintense on DWI. CT identifies cholesteatoma complications (bony erosion, intracranial extension) but cannot characterise the soft tissue nature of the lesion.

1.2 Intrinsic Limitations of the Generic Protocol

Acoustic neuroma size detection limit: the generic protocol detects vestibular schwannomas ≥ 2 mm within the IAC. Sub-2 mm intracanalicular tumours may be below the detection threshold of the 3D CISS even at 0.5–0.7 mm resolution. For surveillance of treated or known small tumours, the high-resolution post-contrast T1 is critical.

Endolymphatic hydrops (Ménière’s disease): conventional MRI with standard gadolinium dosing does not reliably detect endolymphatic hydrops. Dedicated endolymphatic hydrops protocols (intravenous or intratympanic gadolinium, delayed 4-hour imaging) are required and are not part of the generic protocol. See dedicated Ménière’s disease child page.

Perilymph fistula: MRI is not reliable for direct perilymph fistula detection. CT in the dependent position and clinical assessment are the primary tools.

Cochlear patency for cochlear implant planning: the generic protocol provides structural cochlear assessment. Complete cochlear implant candidacy evaluation requires dedicated oblique sagittal sections through the cochlea to confirm modiolus integrity and cochlear nerve presence — covered in the cochlear implant child page.

When dedicated child protocols are required: vestibular schwannoma staging and surveillance; Ménière’s disease (endolymphatic hydrops protocol); cochlear implant candidacy; congenital sensorineural hearing loss (SNHL) workup in children; facial nerve palsy (dedicated full course CN VII); CPA meningioma surgical planning; post-treatment surveillance; cholesteatoma primary and recurrence; sudden SNHL.

2. Main Clinical Indications

2.1 Standard Indications

Asymmetric sensorineural hearing loss (SNHL) is the most common indication for CPA/inner ear MRI. Asymmetric SNHL (≥ 10 dB difference at two or more frequencies, or ≥ 15 dB at any single frequency) triggers imaging because approximately 1–2% of such patients have a retrocochlear cause (predominantly vestibular schwannoma). The generic bilateral protocol is appropriate. When audiological screening with auditory brainstem response (ABR) is abnormal, MRI with bilateral IAC evaluation is the standard of care [1].

Sudden sensorineural hearing loss (sudden SNHL) requires bilateral IAC MRI to exclude: retrocochlear mass (vestibular schwannoma in 3–5%); labyrinthine infarction (T2 signal in the cochlea; DWI restriction); viral labyrinthitis (enhancement of the membranous labyrinth). The generic protocol with contrast is appropriate. Sudden SNHL is a clinical emergency; imaging should be obtained within days of onset for optimal management.

Tinnitus — pulsatile tinnitus requires vascular imaging (CTA or MRA) as the primary modality; non-pulsatile tinnitus with asymmetric features warrants CPA/inner ear MRI to exclude retrocochlear pathology. The generic protocol is appropriate for the retrocochlear survey component.

Vertigo — investigation of central vs peripheral vertigo. MRI is indicated when: vertigo is unilateral; there are associated neurological symptoms; the vestibular pattern is atypical; or central pathology is clinically suspected. The CPA/inner ear protocol assesses the inner ear and vestibular nerve. Associated posterior fossa MRI (for central causes: MS, cerebellar lesions, vascular) may be required simultaneously.

Facial nerve palsy — Bell’s palsy is the most common cause; MRI is not required in typical presentations but is indicated for: atypical or recurrent Bell’s palsy; suspected perineural tumour spread; facial nerve schwannoma; traumatic facial palsy. Post-contrast T1 through the facial canal from the IAC to the stylomastoid foramen is the key sequence.

CPA mass on prior imaging or clinical suspicion — evaluation of suspected or incidentally identified CPA masses. The generic protocol characterises the mass and defines its relationship to the IAC, nerves, and brainstem.

Pre-cochlear implant assessment — the generic protocol provides a baseline structural assessment. Complete cochlear implant candidacy workup requires supplementary sequences (oblique sagittal cochlear nerve section) — see child page.

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

Cochlear implants: standard cochlear implants are a contraindication to MRI at ≥ 1.5T unless the implant is specifically approved for MRI (magnet within the implant must be removed, re-implanted, or the device is certified as MR-conditional at 1.5T). The referring clinician must provide the cochlear implant model and its MR compatibility status before the examination. Some MR-conditional cochlear implants require a head bandage during imaging to prevent magnet dislocation. Verify with the manufacturer and the local cochlear implant team before proceeding.

Hearing aids: remove before the examination. All types of hearing aids (in-ear, behind-ear, bone-anchored) must be removed. Bone-anchored hearing aids (BAHA) with external processor: remove the processor; the titanium osseointegrated post itself is MR-compatible but its proximity to the imaging volume may produce minor susceptibility artefact.

Jewellery and metal near the ear and temporal region: earrings, tragus piercings, helix piercings, industrial piercings, and any metallic objects near the ear must be removed. They produce susceptibility artefacts that extend into the IAC and labyrinth — the exact anatomical target.

Middle ear prostheses (ossicular chain reconstruction): titanium prostheses (TORP, PORP) are MR-compatible and do not require special management. Older stainless steel stapedial prostheses are ferromagnetic and may be displaced; verify the prosthesis type if available. If unknown, 0.55T or 1.5T may reduce risk compared with 3T.

Gadolinium timing for Ménière’s/endolymphatic hydrops protocols: if the examination is specifically for endolymphatic hydrops assessment (Ménière’s disease), the gadolinium must be injected 4 hours before the delayed inner ear imaging sequences. Standard single-dose (0.1 mmol/kg) intravenous gadolinium is followed by a 4-hour wait. This timing requirement is distinct from all other CPA/inner ear indications and must be communicated at the time of booking.

3.2 Patient Positioning on the MRI System

Position: supine, head-first, standard brain MRI position. Head in the head coil. The position is identical to any standard brain MRI.

Coil selection: a dedicated head coil (16–32 channel) is mandatory. The high channel count provides the SNR required for sub-millimetre 3D CISS resolution at 3T. Single-channel or low-channel coils cannot achieve diagnostic sub-millimetre resolution for the inner ear at 1.5T or 3T. The head coil is centred on the temporal bones bilaterally — isocentre at the level of the external auditory canals, approximately at the level of the Frankfurt horizontal plane.

Centring: isocentre at the level of the external auditory canals (approximately the level of the tragus). This centres B0 homogeneity on the temporal bones, optimising shim for the critical 3D CISS acquisition. Off-isocentre positioning (e.g., isocentre at the top of the brain) will degrade B0 homogeneity at the temporal bone level and impair 3D CISS bSSFP performance at 3T.

Head symmetry: verify that the head is not rotated — both temporal bones should appear at the same distance from the midline on the axial localiser. Rotation produces asymmetric B0 distribution and asymmetric coil sensitivity, making bilateral comparison unreliable. A chin-up position aligns the orbitomeatal line with the scanner axis and is standard.

Immobilisation: standard head coil cushions. The 3D CISS acquisition is highly sensitive to motion — even 1–2 mm head motion during the 5–8 minute acquisition degrades the sub-millimetre resolution. Patient instructions and positioning foam are important. For patients with movement disorders or severe tinnitus-related distress, sedation should be considered.

4. Standard Protocol Design

4.1 Mandatory Core Sequences

4.2 Conditional Sequences

4.3 Rationale Summary Per Sequence

3D Heavily T2-Weighted (CISS/DRIVE/FIESTA-C) — the defining sequence of this protocol. This sequence uses the maximum available fluid-to-tissue contrast to simultaneously visualise: - The membranous labyrinth (cochlea, semicircular canals, vestibule) as very bright fluid structures within the dark bony labyrinth - The individual nerve fascicles within the IAC as intermediate-dark structures surrounded by bright CSF - The CPA cistern and its contents - The facial nerve (CN VII) and the vestibulo-cochlear nerve (CN VIII) at their points of entry into the IAC

The sequence is a 3D balanced steady-state free precession (bSSFP) sequence (CISS/FIESTA-C) or a 3D heavily T2-weighted TSE variant (DRIVE, Philips). The bSSFP readout provides maximum T2/T1 weighting at very short TE and TR, producing the highest achievable CSF-to-tissue contrast at clinical field strengths. See Section 10 for the full technical rationale.

Key diagnostic functions of 3D CISS: - Detecting and sizing intracanalicular vestibular schwannomas (appears as filling defect in CSF-bright IAC) - Detecting nerve aplasia or hypoplasia (absent or thin nerve in the IAC — critical for cochlear implant candidacy) - Detecting labyrinthine ossification (T2-dark replacement of the normally bright cochlear turns) - Demonstrating intralabyrinthine schwannoma (filling defect within the cochlea or semicircular canal) - Characterising CPA mass morphology (vestibular schwannoma: intracanalicular T2-dark filling defect; epidermoid: T2-very bright, DWI restricts; meningioma: broad-based, dural tail)

Limitations: susceptibility artefact from air in the middle ear cleft may degrade the CISS image quality at the oval and round windows. At 3T, bSSFP banding artefacts can transiently cross the temporal bone region and must be managed by frequency offset adjustment. Motion degrades the sub-millimetre resolution — a blurred CISS is non-diagnostic.

T1-weighted TSE axial provides: - Baseline T1 map before contrast - T1-bright lesions: rare intralabyrinthine haemorrhage; some schwannomas have intrinsic T1 signal; lipomas (T1-bright, fat-saturated away) of the IAC (rare) - Reference for post-contrast comparison - Bone marrow signal assessment (petrous apex marrow; cholesterol granuloma of the petrous apex is T1-bright)

Cholesterol granuloma: the petrous apex T1-bright lesion most commonly confused with other pathology. Its T1 hyperintensity on pre-contrast T1 (and maintained T1 brightness post-contrast — it does not enhance like a solid tumour but remains bright) is the key diagnostic feature. The T1 pre-contrast is mandatory for this reason.

T2-weighted TSE axial provides standard T2 contrast for: - Brain parenchyma adjacent to the CPA (posterior fossa assessment) - CPA mass T2 signal (vestibular schwannoma: heterogeneously T2-bright to intermediate; epidermoid: T2-very bright; meningioma: T2-isointense to brain) - Any associated posterior fossa abnormality (cerebellar signal, fourth ventricular compression) - Complements the 3D CISS for overall spatial orientation

DWI and ADC — mandatory because cholesteatoma is the most important differential diagnosis requiring DWI: - Cholesteatoma: b=1000 hyperintense; ADC approximately 0.4–0.7 × 10⁻³ mm²/s (restricted diffusion from the keratinous desquamated epithelial debris) - Epidermoid cyst (CPA): b=1000 hyperintense (“shining” on DWI — the most reliable sign); ADC low - Arachnoid cyst: b=1000 not restricted; same signal as CSF - Vestibular schwannoma: b=1000 variable; usually not significantly restricted

The CPA DWI distinction between epidermoid and arachnoid cyst is one of the highest-confidence DWI applications in neuroradiology: DWI is pathognomonic when the CISS/T2 shows a CPA mass that follows CSF signal but DWI shows restriction — this virtually confirms epidermoid over arachnoid cyst.

Post-contrast T1 fat-suppressed — mandatory when contrast is administered: - Enhancing intracanalicular vestibular schwannoma (smallest detectable is approximately 2 mm with thin-slice post-contrast T1) - Labyrinthitis (enhancement of the membranous labyrinth) - Facial neuritis (Bell’s palsy: enhancement of the geniculate ganglion and tympanic segment) - Leptomeningeal carcinomatosis (enhancement of cranial nerve sheaths in the CPA) - Vascular malformation enhancement - Post-contrast subtraction (post minus pre T1) can improve small enhancing lesion conspicuity

Fat suppression: SPAIR or Dixon at isocentre (temporal bones are at B0 centre with correct head coil positioning). STIR is not used post-contrast.

4.4 Sequence Matching and Cross-Sequence Consistency

The 3D CISS acquisition is the foundational dataset from which all reformats are generated. Its axial isotropic volume is reformatted into: - Coronal plane: optimal for visualising the course of CN VII and CN VIII within the IAC - Sagittal oblique perpendicular to IAC: optimal for identifying individual nerves (cochlear nerve anterior-inferior; facial nerve anterior-superior; superior vestibular nerve posterior-superior; inferior vestibular nerve posterior-inferior — the classic “four quadrant” IAC anatomy)

The pre-contrast T1 and post-contrast T1 must use identical geometry for accurate subtraction and side-by-side comparison. Any geometry change between pre and post contrast renders the subtraction unreliable.

The DWI must cover the same volume as the 3D CISS to enable co-registration of DWI-positive findings with their anatomical CISS correlate.

4.5 Fat Suppression

Fat suppression is required only for the post-contrast T1 sequences. At isocentre with a head coil, SPAIR provides reliable fat suppression. Dixon is the preferred alternative at 3T for its B0-independence.

The pre-contrast T1 is acquired without fat suppression to maintain T1 contrast for: (a) T1-bright lesions (cholesterol granuloma, lipoma, haemorrhage, high-protein cyst); (b) baseline fat signal for comparison with post-contrast fat-suppressed images.

3D CISS does not use fat suppression (bSSFP contrast mechanism; fat suppression would alter the steady-state condition).

4.6 Slice Positioning — Complete Technical Reference

Why Precise Positioning Is Critical

The CPA and inner ear occupy a small volume (approximately 3 × 3 × 2 cm per side) at a specific location within the posterior fossa. The critical structures (cochlear turns, IAC nerves) are oriented obliquely relative to the standard scanner axes. The standard axial plane of the brain is acceptable for most CPA assessment, but specific manoeuvres — the coronal reformat along the IAC axis, and the sagittal oblique perpendicular to the IAC — are essential for nerve identification and require that the 3D CISS be isotropic (so that these reformats are full-resolution).

Anatomical Landmarks

Internal auditory canal (IAC): the bony canal connecting the posterior cranial fossa to the inner ear. Approximately 8–10 mm long; runs laterally and slightly anteriorly from the CPA to the fundus (where the nerves exit to the cochlea, semicircular canals, and facial canal).

IAC fundus (lateral end): where the IAC opens into the labyrinthine compartments. The crista falciformis (horizontal bone crest) divides the fundus into superior (facial nerve + superior vestibular nerve) and inferior (cochlear nerve + inferior vestibular nerve) compartments. The transverse crest (Bill’s bar) divides the superior compartment vertically (facial nerve anteriorly).

CPA cistern: the CSF-filled angle between the cerebellum, pons, and petrous bone. The CPA tumours typically originate in the IAC and extend into the CPA cistern.

Porus acusticus: the medial opening of the IAC into the CPA cistern; the site where the CN VII/VIII bundle enters the IAC.

Planning Sequence

1. Three-plane localiser (standard brain)
2. From the axial localiser: identify the IACs bilaterally; plan the 3D CISS slab to encompass both IACs, both labyrinths, and the CPA cisterns (typically a 5 × 5 cm slab centred on the temporal bones)
3. Confirm that the 3D CISS slab includes: the porus acusticus (medial) through the fundus (lateral) bilaterally; the cochlea; the semicircular canals; the CPA cisterns
4. Post-acquisition reformats: coronal and sagittal oblique are generated from the isotropic dataset

Axial Planning (3D CISS primary acquisition)

Orientation: standard axial (parallel to the orbitomeatal line or equivalent). The primary 3D CISS acquisition is performed in the axial plane at isotropic resolution (0.5–0.7 mm). The entire petrous bone bilaterally is covered in a single 3D slab.

Coverage: from the inferior margin of the petrous bone (cochlea aqueduct level) to the superior margin of the petrous bone (superior semicircular canal level). Craniocaudal extent approximately 3–4 cm for a targeted temporal bone slab; some centres prefer full posterior fossa coverage (10–12 cm) at the cost of slightly longer acquisition time for a thicker 3D slab.

Lateral coverage: extends from mastoid to cochlea bilaterally, plus the CPA cisterns medially to the pons.

Phase encoding direction: A-P for axial 3D CISS. This displaces any EPI-equivalent ghost or motion artefact away from the temporal bones.

Coronal Reformat from 3D CISS

The coronal reformat from the isotropic 3D CISS is generated post-acquisition, perpendicular to the long axis of the IAC. The IAC runs approximately 15–20° oblique to the standard coronal plane. The true IAC-perpendicular coronal plane is obtained by rotating the reformat plane to be perpendicular to the IAC long axis as seen on the axial localiser from the 3D CISS dataset.

Diagnostic value: the coronal reformat provides a face-on view of the IAC, showing the four quadrant division (facial nerve anterosuperior, cochlear nerve anteroinferior, superior vestibular nerve posterosuperior, inferior vestibular nerve posteroinferior) — the only reliable way to identify individual nerves within the IAC on clinical MRI.

Sagittal Oblique Reformat (Perpendicular to IAC)

A sagittal oblique plane perpendicular to the IAC long axis produces cross-sectional “axial cuts” through the IAC. At each level along the IAC, the four nerves can be individually identified. This is the standard projection used for cochlear nerve identification in cochlear implant candidacy assessment.

Axial Planning for the T1 and T2 Sequences

The T1 and T2 axial sequences are planned to cover the posterior fossa from the brainstem to the temporal bones, using the standard axial brain acquisition. These sequences do not require targeted temporal bone positioning — they provide the CPA context and posterior fossa assessment.

Section 4.6 Dedicated Bibliography

Bartindale MR, et al. The anatomy of the posterior fossa: implications for magnetic resonance imaging. Neuroimaging Clin N Am. 2016;26(3):355–366. DOI: not verified. (Technical / Foundational) — Anatomical reference for CPA and inner ear MRI planning; documents landmark-based positioning methodology.

Casselman JW, et al. Pathology of the membranous labyrinth: comparison of T1- and T2-weighted and gadolinium-enhanced spin-echo and 3DFT-CISS imaging. AJNR Am J Neuroradiol. 1993;14(1):59–69. PMID: 8427100. (Technical / Foundational) — Original description of 3D CISS for inner ear imaging; establishes the technical requirements for IAC and labyrinthine MRI positioning.

5. Optimisation Strategy

5.1 Artifact Reduction by Source

bSSFP banding artefact (3D CISS at 3T): the most important quality issue specific to this protocol. As documented in the 2D vs 3D child page, bSSFP sequences are sensitive to B0 off-resonance. At 3T, the banding artefact in 3D CISS appears as alternating dark bands that can cross the inner ear at any position, depending on local B0. Mitigation: (a) optimise B0 shimming on the temporal bones before the 3D CISS acquisition; (b) adjust the centre frequency in 100–200 Hz steps to shift the banding node away from the IAC/cochlea (described in Section 8.3); (c) on systems with CISS (double acquisition): the CISS magnitude combination of two bSSFP acquisitions with different RF phases effectively suppresses banding — do not modify this CISS-specific protocol feature. At 1.5T, banding is less severe; at 0.55T, 3D CISS is generally not feasible at sub-millimetre resolution.

Air-tissue susceptibility at the middle ear and mastoid: the air-filled middle ear space and mastoid air cells produce local B0 disturbances adjacent to the labyrinth. The cochlear windows (oval and round window) are particularly susceptible to signal dropout from this air-bone-soft tissue boundary. Mitigation: accept this as an inherent limitation; the inner ear structures are not within the air space and are largely unaffected; the windows region is the only area of concern.

Motion artefact: the 3D CISS acquisition at sub-millimetre isotropic resolution over 5–8 minutes is extremely sensitive to patient head motion. Even 1–2 mm motion during the acquisition blurs the image to the point of non-diagnosticity — the fine nerve structures (< 1 mm) disappear. Mitigation: patient instruction; head immobilisation foam; consider sedation for patients with movement disorders; use the shortest feasible 3D CISS acquisition time (CS acceleration at 3T can reduce to 3–4 minutes).

Dental metalwork susceptibility: as noted in the neck and brachial plexus protocols, posterior dental metalwork produces susceptibility artefacts that extend several centimetres. For the IAC and inner ear (located at the petrous apex, well posterior to the mandibular condyle), dental metalwork rarely affects the diagnostic region directly. It may affect the anterior CPA on the side of the implant. Document if present.

Ghost from CSF pulsation (CPA cistern): CSF pulsation in the CPA cistern produces artefacts in the phase-encoding direction. Phase direction A-P for axial sequences displaces these artefacts anteroposteriorly; pulsation-related artefacts crossing the IAC (the target) are minimised with A-P phase direction.

5.2 Protocol Efficiency and Throughput

A complete CPA/inner ear MRI — T1 axial + T2 axial + 3D CISS + DWI + post-contrast T1 axial + post-contrast T1 coronal — requires approximately 25–35 minutes at 3T.

An abbreviated protocol — 3D CISS + DWI + post-contrast T1 axial — covers the essential diagnostic requirements in approximately 15–20 minutes. This is adequate for routine vestibular schwannoma detection screening. The T1 and T2 provide supporting information but do not change management in most cases.

Compressed sensing 3D CISS (available on Siemens, Philips, and GE) can reduce the 3D CISS from 6–8 minutes to 3–4 minutes at equivalent or near-equivalent quality — a significant efficiency gain for a protocol dominated by one long sequence.

5.3 Field Strength Considerations

3T is strongly preferred for CPA/inner ear MRI because: - The sub-millimetre 3D CISS (0.5–0.7 mm) requires the higher SNR of 3T to achieve diagnostic quality within acceptable scan time - At 1.5T, the 3D CISS at 0.7 mm isotropic is achievable but with longer acquisition times (8–12 minutes) - At 3T, the 3D CISS can be acquired at 0.5 mm isotropic in 5–7 minutes with GRAPPA R=2

1.5T is still clinically adequate for most CPA/inner ear indications: - Vestibular schwannoma ≥ 4 mm: detectable at 1.5T - Labyrinthine obliteration: detectable - DWI for cholesteatoma/epidermoid: adequate at 1.5T

Limitation at 1.5T: small (< 3 mm) intracanalicular vestibular schwannomas and detailed cochlear nerve assessment for cochlear implant candidacy are at the limit of 1.5T 3D CISS capability. Expert centres with this specific requirement should operate at 3T.

bSSFP banding at 3T: more severe than at 1.5T (see Section 5.1). The standard mitigation (CISS double acquisition) is product-implemented and does not require technologist intervention on Siemens CISS; on GE FIESTA-C, banding management requires centre frequency adjustment.

SAR at 3T: the 3D CISS (bSSFP) has relatively low SAR at 3T due to the low flip angle typically used (25–35°). CISS SAR is rarely a limiting factor at 3T even with multiple REST slabs or additional sequences.

6. Contrast Use Principles Specific to CPA/Inner Ear MRI

6.1 Non-Contrast Standard Protocol — Sufficient For

Non-contrast CPA/inner ear MRI (3D CISS + T1 + T2 + DWI) is diagnostically adequate for: - Initial screening for vestibular schwannoma when 3D CISS detects or excludes an intracanalicular filling defect (a normal 3D CISS with no IAC filling defect effectively excludes vestibular schwannoma ≥ 3 mm in experienced centres) - Epidermoid cyst characterisation (DWI restriction + T2-bright = diagnostic) - Labyrinthine ossification (T2 signal loss in the cochlea on CISS) - Congenital inner ear malformation assessment - Pre-cochlear implant structural assessment

Non-contrast CISS alone as a screening tool: several published series [2, 3] have examined whether 3D CISS alone (without post-contrast T1) is adequate for vestibular schwannoma screening. The sensitivity of 3D CISS for tumours > 2–3 mm is very high (> 95%). For tumours < 2 mm, post-contrast T1 is more sensitive. Current practice at most expert centres uses CISS + contrast for all clinical indications where a schwannoma must be confidently excluded [1].

6.2 Gadolinium Indicated — Region-Specific Contexts

Post-contrast T1 is required for: - Any clinical indication where vestibular schwannoma detection is the primary question (asymmetric SNHL, unilateral tinnitus, vertigo with retrocochlear suspicion): contrast detects small (< 3 mm) enhancing intracanalicular tumours not visible on CISS - Facial nerve pathology (Bell’s palsy, suspected schwannoma, perineural spread): enhancement of the geniculate ganglion and intratemporal facial nerve segments - Labyrinthitis: enhancement of the membranous labyrinth (cochlea, semicircular canals) - Leptomeningeal carcinomatosis: enhancement of CPA cranial nerve sheaths - Post-treatment assessment (schwannoma after radiosurgery): enhancement pattern and volume change

6.3 Post-Contrast Acquisition Timing

Standard post-contrast T1-FS sequences are acquired at 3–5 minutes post-injection (equilibrium phase). No special arterial or delayed phase timing is required for the standard CPA/inner ear protocol.

Endolymphatic hydrops (Ménière’s disease) protocol exception: intravenous gadolinium injected 4 hours before imaging; double or triple dose (0.2–0.3 mmol/kg); the delayed acquisition allows gadolinium to distribute into the perilymph space but not the endolymph space. This timing completely changes the examination schedule and must be pre-planned.

7. Reporting Essentials

7.1 Interpretation Framework

CPA/inner ear MRI reporting follows a systematic bilateral evaluation:

For each IAC (bilateral): - IAC morphology: width, length, any asymmetry - Nerve bundle: four quadrant anatomy on sagittal oblique reformat; is the cochlear nerve identifiable? - Any filling defect within the IAC on 3D CISS - Enhancement on post-contrast T1

For each labyrinth (bilateral): - Cochlear T2 signal (CISS): normal bright fluid vs dark = obliteration; focal dark = intralabyrinthine schwannoma - Semicircular canals: visible / not visible (obliteration) - Vestibule: normal

CPA cisterns (bilateral): - Any mass: location, T2 signal, DWI, size, enhancement - Relationship to IAC (intracanalicular extension) - Relationship to brainstem and cerebellum - CPA cistern symmetry

Brainstem and posterior fossa: - T2 signal (for demyelination, vascular, mass) - Fourth ventricle: normal / compressed / displaced

Differential axis for CPA masses: - Intracanalicular origin + CISS filling defect + enhancement → vestibular schwannoma - Broad dural base + no intracanalicular extension + isointense to brain → meningioma - T2-very bright + DWI restriction + no enhancement → epidermoid cyst - T2-very bright + no DWI restriction + no enhancement → arachnoid cyst - CPA + pulsatile lesion + TOF bright → vascular lesion (paraganglioma, vascular loop)

7.2 Mandatory Reporting Checklist

Technical quality: - [ ] Field strength; coil - [ ] 3D CISS quality: banding artefact present? (location — affects IAC? Yes/No); motion blurring? - [ ] DWI: b=1000 and ADC available - [ ] Contrast: if given — agent, dose, timing

Right IAC: - [ ] IAC dimensions (mm): width; length - [ ] CISS filling defect: absent / present (size in mm; intracanalicular vs CPA component) - [ ] Cochlear nerve: present (normal calibre) / thin / absent - [ ] Post-contrast T1: no enhancement / enhancement (size, location)

Left IAC: same checklist

Bilateral labyrinth: - [ ] Right cochlea CISS signal: normal / reduced (obliteration risk — specify cochlear turns affected) - [ ] Left cochlea CISS signal - [ ] Semicircular canals: normal bilateral / abnormal (specify side and canal)

CPA cisterns: - [ ] Mass: absent / present (bilateral note if applicable) - [ ] DWI: no restriction / restriction (location, lesion)

Posterior fossa: - [ ] Brainstem T2: normal / abnormal - [ ] Cerebellum: normal / abnormal - [ ] Fourth ventricle: normal / obstructed

7.3 Structured Reporting

Reports must include: Indication (asymmetric SNHL / tinnitus / vestibular schwannoma surveillance / other); Technique (field strength, coil, sequences, 3D CISS voxel size, contrast: agent, dose); Comparison (prior MRI date); Findings (bilateral systematic review as per checklist); Impression (primary differential; vestibular schwannoma: measurement in three dimensions; labyrinthine findings; DWI findings); Limitations (CISS banding artefact at specific level; motion; metal artefact).

7.4 Incidental Findings — Clinical Decision Framework

Usually benign: asymmetric pneumatisation of mastoid and petrous apex (normal variation); mild asymmetry of IAC width (up to 2 mm asymmetry is within normal limits); incidental small arachnoid cyst in the CPA (T2-bright, no DWI restriction, no enhancement — stable; report for documentation only).

May require follow-up: asymmetric enhancement of the IAC without a CISS filling defect (borderline finding — suggest short-term follow-up MRI in 12 months or audiological review); petrous apex lesion with intermediate signal characteristics (cholesterol granuloma vs mucocele vs asymmetric marrow fat — further characterisation with CT may help).

Urgent communication required: large CPA mass with fourth ventricular compression or brainstem displacement; unexpected leptomeningeal enhancement (carcinomatosis — new cancer diagnosis); unexpected brainstem infarction; large endolymphatic sac tumour in a VHL disease context.

8. MRI Technologist Pearls

8.1 Sequence Order Logic

1. Three-plane localiser ← standard brain positioning
2. 3D CISS ← always first; most critical sequence; motion-sensitive at sub-mm resolution; best done when patient is fresh and compliant
3. T2 TSE axial ← structural overview
4. T1 TSE axial ← pre-contrast reference; T1-bright lesions
5. DWI ← cholesteatoma/epidermoid detection
6. Contrast injection
7. Post-contrast T1-FS axial ← main enhancement detection
8. Post-contrast T1-FS coronal ← facial nerve course; optional

The 3D CISS is placed first because: (a) it is the most diagnostically critical sequence; (b) it is the most motion-sensitive; (c) if the patient cannot complete the examination, having the CISS provides the primary diagnostic information.

8.2 Positioning Tricks

Verify isocentre at the tragus level: for CPA/inner ear MRI, the standard brain isocentre (level of the nasion or midface) is too superior. Lower the isocentre to the level of the tragus (external auditory canal). This places the temporal bones at the B0 field centre — essential for 3D CISS bSSFP quality at 3T.

bSSFP banding management before starting the 3D CISS: run a B0 map or use the 3D CISS test scan function to visualise where banding crosses the FOV. If a band crosses the IAC/cochlea, adjust the centre frequency in small steps (50–100 Hz) to shift the band away. Do not start the full diagnostic 3D CISS with a band crossing the target anatomy.

For CISS at 3T with persistent banding: if banding cannot be fully suppressed by shimming and centre frequency adjustment, the DRIVE (Philips) or 3D SPACE T2 alternative provides banding-free inner ear imaging at a slight SNR cost and slightly longer acquisition. DRIVE is specifically designed to avoid bSSFP banding while maintaining high T2 fluid contrast.

8.3 Fast Salvage Protocol

Approximately 10–11 minutes — covers the three primary diagnostic questions for the most common indications (vestibular schwannoma, epidermoid/cholesteatoma).

8.4 Common Avoidable Errors

9. Quality Control Checklist

• Isocentre verified at tragus level (external auditory canal)
• 3D CISS: banding artefact location assessed — does not cross IAC or cochlea bilaterally
• 3D CISS: motion assessment — IAC nerves sharply defined (sub-mm structure visible)
• 3D CISS: both cochlear turns visible bilaterally (bright fluid signal in all turns)
• 3D CISS: coronal and sagittal oblique reformats generated and reviewed
• DWI: b=1000 and ADC map generated
• T1 pre-contrast: petrous apex signal documented for cholesterol granuloma exclusion
• Post-contrast T1-FS: fat suppression uniform at temporal bone level
• Post-contrast T1-FS: both IAC fundus regions covered
• If contrast given: injection time documented; timing 3–5 min post-injection
• Bilateral coverage confirmed: both IACs, both labyrinths, both CPA cisterns
• Posterior fossa coverage adequate (brainstem, cerebellum visible)

10. Advanced Technical Parameters

10.1 3D Heavily T2-Weighted (CISS / FIESTA-C / DRIVE)

Tissue Contrast Logic

The 3D bSSFP (CISS, FIESTA-C) produces maximum fluid-to-tissue contrast via the T2/T1 ratio mechanism. In bSSFP steady state, the signal of a tissue is proportional to:

S_bSSFP ∝ (T2/T1) × M₀

For CSF: T2/T1 ≈ 2000/4000 = 0.5 For neural tissue: T2/T1 ≈ 80/900 = 0.09 Ratio CSF/neural tissue ≈ 5.6 — the CSF appears approximately 5× brighter than neural tissue.

This CSF-to-tissue contrast is maximised at very short TE/TR (maintaining the bSSFP steady state) and at the flip angle near 30–40°.

The DRIVE (Philips) variant uses a heavily T2-weighted RARE readout (very long effective TE) rather than bSSFP. DRIVE provides T2-based fluid-bright contrast that is less susceptible to bSSFP banding but has slightly lower fluid-to-tissue contrast ratio.

Key Parameters

Vendor equivalents: - Siemens: CISS (double-acquisition bSSFP with magnitude reconstruction to suppress banding) - GE: FIESTA-C (CISS equivalent; C = combined, for banding suppression) - Philips: DRIVE (RARE-based; banding-free alternative) or bSSFP - Canon: FASE (equivalent)

CISS vs FIESTA-C vs DRIVE: - Siemens CISS and GE FIESTA-C: double bSSFP acquisition with different RF phase increments; magnitude combination suppresses banding → two acquisitions slightly longer but banding-free - Philips DRIVE: single acquisition; no banding; slightly lower fluid-to-tissue contrast than bSSFP at equivalent voxel size; recommended at 3T when bSSFP banding is problematic

Section 10 Dedicated Bibliography

Casselman JW, et al. Pathology of the membranous labyrinth: comparison of T1- and T2-weighted and gadolinium-enhanced spin-echo and 3DFT-CISS imaging. AJNR Am J Neuroradiol. 1993;14(1):59–69. PMID: 8427100. (Technical / Foundational) Original 3D CISS methodology for inner ear MRI; the primary technical reference for 3D bSSFP inner ear imaging.

Naganawa S, et al. Preferred parameters for 3D-FLAIR of the endolymphatic space using 3T MRI. Magn Reson Med Sci. 2012;11(4):259–264. PMID: 23257854. DOI: 10.2463/mrms.11.259. (Technical / Foundational) Technical reference for inner ear MRI sequences; 3T parameter optimisation for inner ear protocols.

11. Evidence Gaps and Ongoing Debate

Non-contrast 3D CISS alone vs CISS + contrast for vestibular schwannoma screening: multiple studies have shown that 3D CISS alone detects vestibular schwannomas ≥ 3 mm with sensitivity > 95%. For tumours < 2–3 mm, post-contrast T1 is substantially more sensitive. Whether the added diagnostic yield of post-contrast T1 over 3D CISS alone justifies routine gadolinium use in all asymmetric SNHL patients is debated. Some guidelines suggest contrast can be omitted when 3D CISS shows no filling defect in the IAC; others mandate contrast for all IAC imaging. The threshold at which contrast is mandatory remains centre-dependent [1, 2, 3].

Optimal voxel size for 3D CISS: while 0.5–0.7 mm isotropic is the current expert consensus standard, formal studies comparing detection rates at different voxel sizes (0.5 vs 0.7 vs 0.9 mm) for small vestibular schwannomas are limited. The lower bound of meaningful resolution for cochlear nerve identification in the context of cochlear implant candidacy is not formally established.

CISS vs DRIVE for 3T CPA imaging: CISS and DRIVE provide different trade-offs at 3T (CISS: higher fluid contrast, banding risk; DRIVE: lower fluid contrast, no banding). No prospective comparative study has formally established superiority for clinical diagnosis across all indications. Centre choice depends on scanner platform and protocol preference.

Endolymphatic hydrops imaging standardisation: the gadolinium-enhanced delayed inner ear imaging protocol for Ménière’s disease (intravenous vs intratympanic gadolinium; standard vs double dose; 4h vs 24h delay; CISS vs FLAIR post-gadolinium) remains unstandardised across centres. No international guideline has specified a single reference protocol.

AI-assisted inner ear structure segmentation: automated cochlea and semicircular canal segmentation from 3D CISS data has been demonstrated in research settings for surgical planning and pre-cochlear implant assessment. No clinically validated, regulatory-cleared tool exists at the time of writing.

12. Evidence-Based References

A. Guidelines / Consensus / Society Recommendations

[1] Kessler MM, et al. ACR Appropriateness Criteria Hearing Loss and/or Vertigo. J Am Coll Radiol. 2017;14(5S):S174–S180. PMID: 28473086. DOI: 10.1016/j.jacr.2017.01.040. (High — Practice guideline) ACR appropriateness criteria for hearing loss and vertigo; designates MRI with gadolinium as usually appropriate for asymmetric SNHL evaluation; primary indication framework for CPA/inner ear MRI.

B. Systematic Reviews / Meta-analyses

[2] Fortnum H, et al. Imaging of sporadic unilateral acoustic neuroma: results of a systematic review. Br J Radiol. 2009;82(978):572–582. PMID: 19482958. DOI: 10.1259/bjr/10654546. (High — Systematic review) Systematic review of MRI for vestibular schwannoma detection; addresses sensitivity of contrast vs non-contrast protocols for intracanalicular tumours.

[3] Stuckey SL, Harris MR, Mannolini SM. Detection of acoustic schwannoma: use of constructive interference in the steady state three-dimensional MR. AJNR Am J Neuroradiol. 1996;17(7):1219–1225. PMID: 8873651. (Moderate — Prospective study) Early clinical validation of 3D CISS for vestibular schwannoma detection; establishes the diagnostic role of the 3D CISS sequence for IAC masses.

C. Important Prospective / Original Studies

[4] Vernooij MW, et al. Incidental findings on brain MRI in the general population. N Engl J Med. 2007;357(18):1821–1828. PMID: 17978290. DOI: 10.1056/NEJMoa070972. (High — Population study) Provides context for incidental CPA findings; documents prevalence of incidental inner ear and CPA findings.

D. Technical MRI Papers

[5] Casselman JW, et al. Pathology of the membranous labyrinth: comparison of T1- and T2-weighted and gadolinium-enhanced spin-echo and 3DFT-CISS imaging. AJNR Am J Neuroradiol. 1993;14(1):59–69. PMID: 8427100. (Technical / Foundational) Original 3D CISS methodology for inner ear MRI; the primary technical reference establishing CISS as the standard inner ear sequence.

[6] Naganawa S, et al. Preferred parameters for 3D-FLAIR of the endolymphatic space using 3T MRI. Magn Reson Med Sci. 2012;11(4):259–264. PMID: 23257854. DOI: 10.2463/mrms.11.259. (Technical / Foundational) Inner ear protocol parameters at 3T; documents technical requirements for endolymphatic space MRI including sequence parameter optimisation.

End of document — MRI Cerebellopontine Angle and Inner Ear Generic Standard Protocol — MRIninja v1.0 — May 2026 This master page is the reference for all future CPA/inner ear child pages including: vestibular schwannoma staging and surveillance; Ménière’s disease / endolymphatic hydrops protocol; pre-cochlear implant assessment; congenital SNHL in children; sudden SNHL; facial nerve palsy; cholesteatoma primary and recurrence; epidermoid vs arachnoid cyst.