MRI Endolymphatic Hydrops — Generic Standard Protocol
Required Protocol at a Glance
Mandatory core sequences for this examination. Detailed rationale, conditional additions and optimisation notes are provided later in the protocol.
MRIninja Knowledge Base | Master / General Protocol Page Related pages: MRI CPA and Inner Ear Generic Standard Protocol · MRI Middle Ear and Petrous Bone Generic Standard Protocol Version 1.0 — May 2026
1. Executive Summary
The MRI protocol for endolymphatic hydrops (EH) is operationally, technically, and pharmacokinetically unlike any other MRI protocol in clinical practice. Its fundamental principle — injecting gadolinium 4 hours before imaging rather than minutes before — derives from the unique anatomy of the blood-labyrinth barrier and the pharmacokinetics of gadolinium distribution between the two fluid compartments of the membranous labyrinth. No imaging technique existed for in vivo diagnosis of endolymphatic hydrops before Naganawa et al. [1] introduced delayed gadolinium-enhanced MRI in 2007. Prior to this, EH diagnosis was exclusively clinical (Bárány Society criteria [2]) or post-mortem histological.
The protocol is best understood through its physiology: intravenous gadolinium distributes from the bloodstream into the perilymph (the outer labyrinthine fluid compartment, contiguous with CSF via the cochlear aqueduct) over a period of hours. The endolymph (the inner compartment, enclosed within the membranous labyrinth) is protected by the blood-labyrinth barrier and does not accumulate gadolinium. After approximately 4 hours, the T1 of perilymph is significantly shortened by its gadolinium content, while the T1 of endolymph remains unchanged at the long T1 of pure fluid (~4000 ms at 3T). Dedicated MRI sequences exploit this T1 differential to visualise the two compartments separately — making the expanded endolymph of EH directly measurable and gradable.
This master page covers the complete operational and technical framework for EH MRI, including: the 4-hour delay management; gadolinium dosing strategies; the 3D FLAIR inverted-image technique; the 3D T1 direct enhancement technique; the Naganawa grading system; bilateral assessment; and serial monitoring protocols. Future child pages will address specific disease contexts (definite Ménière’s disease, secondary EH, bilateral EH treatment monitoring, intratympanic gadolinium protocol).
1.1 Core Strengths
The only in vivo EH biomarker: EH MRI with delayed gadolinium is currently the only non-invasive method that directly visualises and grades endolymphatic expansion in the living patient. It provides structural evidence for or against the EH diagnosis that clinical criteria alone cannot provide, and identifies bilateral disease that is clinically silent in 30–40% of Ménière’s patients [3].
Bilateral assessment in a single examination: the protocol images both inner ears simultaneously, detecting bilateral EH even when only one ear is symptomatic. This has direct implications for treatment decisions, particularly for ablative therapies (intratympanic gentamicin, labyrinthectomy) where bilateral disease must be excluded before destructive intervention on the symptomatic side.
Grading of disease severity and distribution: the Naganawa system [1, 4] provides semi-quantitative grades (0–2) for cochlear EH and vestibular EH separately. Cochlear EH predominates in cochlear Ménière’s; vestibular EH (saccular dilatation) is the most characteristic finding in classic Ménière’s disease. The distribution pattern (cochlear-predominant vs vestibular-predominant vs combined) guides audiological and vestibular interpretation.
Serial monitoring capacity: in patients undergoing medical or surgical treatment (intratympanic dexamethasone, endolymphatic sac decompression, intratympanic gentamicin), serial EH MRI provides objective imaging evidence of hydrops response to treatment — a capability unavailable with any other modality.
1.2 Intrinsic Limitations of the Protocol
Operational complexity — the 4-hour delay is the defining constraint: the patient must be injected at a scheduled time and remain available for exactly 4 hours (acceptable range 3.5–4.5 hours) before re-entering the scanner. This requires a dedicated booking infrastructure, patient compliance, and institutional capacity for a two-stage appointment. Unlike any other MRI protocol, the diagnostic value of the examination is entirely dependent on accurate delay timing — an examination performed at 2 hours or 6 hours provides non-diagnostic or unreliable grading.
Gadolinium is mandatory and the dose affects image quality: there is no non-contrast EH MRI alternative. Patients with severe renal impairment (eGFR < 30 mL/min/1.73 m²) or prior severe GBCA reaction cannot undergo this protocol. The gadolinium dose (single, double, or triple) affects the perilymph T1 shortening and therefore the quality of the EH signal — dose standardisation is important for reproducibility.
Disease-state variability: EH severity varies between attack and inter-ictal phases. A single delayed MRI represents a snapshot. Grade 1 at one examination does not preclude grade 2 at another. For clinical decision-making, this variability must be acknowledged and the imaging should ideally be correlated with the patient’s current symptom status.
Resolution at the limit of clinical MRI: the cochlear duct measures 0.5–1.0 mm in width; reliable cochlear EH grading requires the highest achievable in-plane resolution (0.6–0.8 mm isotropic at 3T). At 1.5T or with suboptimal 3D FLAIR parameters, cochlear EH grading may be unreliable, particularly for grade 1 (mild) EH.
TI calibration dependency: the 3D FLAIR post-Gd sequence requires a site-specific TI calibrated for the gadolinium dose and delay interval used at each institution. A standard brain FLAIR TI will null the gadolinium-containing perilymph along with CSF — making the examination non-diagnostic. This calibration requirement is the single most important technical hurdle for departments implementing EH MRI.
2. Main Clinical Indications
2.1 Standard Indications
Definite or probable Ménière’s disease (Bárány Society criteria [2]) is the primary indication. EH MRI provides the imaging biomarker that supports or questions the clinical diagnosis, grades disease severity in both ears, and identifies bilateral involvement. The protocol is not required for diagnosis (which remains clinical) but significantly augments diagnostic certainty and surgical planning.
Clinically suspected Ménière’s disease with diagnostic uncertainty: when the clinical presentation is atypical, incomplete, or when the treating physician requires objective evidence before committing to long-term treatment, EH MRI provides additional information. A clearly negative EH MRI (bilateral grade 0 in both cochlea and vestibule) raises doubt about a Ménière’s diagnosis. A strongly positive bilateral EH MRI may confirm subclinical bilateral disease.
Pre-ablative therapy planning: before intratympanic gentamicin or surgical labyrinthectomy, EH MRI excludes bilateral disease on the side to be treated and confirms the degree of hydrops in the target ear. Operating on the wrong ear or on a patient with bilateral disease carries severe consequences; EH MRI reduces this risk.
Treatment monitoring: serial EH MRI at standardised intervals (typically 6–12 months) after intratympanic dexamethasone, endolymphatic sac decompression, or other interventions provides objective EH grade change data. This is the most reproducibility-demanding use of the protocol — identical dose, delay, and acquisition parameters are mandatory.
Secondary endolymphatic hydrops: EH has been documented in delayed post-traumatic conditions, post-viral labyrinthitis, autoimmune inner ear disease, and superior canal dehiscence. EH MRI confirms the structural substrate in these secondary EH conditions.
Vestibular migraine vs Ménière’s disease differentiation: these conditions share clinical overlap. EH MRI showing positive EH supports Ménière’s; a negative EH study in a suspected Ménière’s patient increases the probability of vestibular migraine or another aetiology. The specificity of EH MRI for this differential is moderate [5].
2.2 Urgent Red Flags
Endolymphatic hydrops and Ménière’s disease are chronic conditions. The EH MRI protocol is elective and scheduled. There are no urgent indications for this specific protocol. Acute presentations with sudden deafness, severe vertigo, or new neurological signs should be evaluated first with the standard CPA/inner ear protocol (see companion page) to exclude: labyrinthine infarction; vestibular schwannoma; retrocochlear mass; Ramsay Hunt syndrome. The EH MRI is arranged subsequently once acute pathology has been excluded.
3. Preparation Reference
Universal MRI safety screening belongs to the general MRI preparation page and is not repeated here. The preparation requirements specific to EH MRI are governed almost entirely by the gadolinium pharmacokinetic protocol.
3.1 Anatomy-Specific and Protocol-Specific Preparation Items
The 4-hour delay is a logistical preparation requirement, not merely a technical one: the referring clinician, the patient, and the radiology department must all understand that this is a two-stage procedure requiring two separate contacts with the department: 1. Gadolinium injection (at the scheduled injection time) 2. MRI scanning (exactly 4 hours later)
The patient must not eat or have other medical procedures between injection and imaging that would interfere with gadolinium pharmacokinetics or MRI compatibility. Light food and water are acceptable. Strenuous exercise is discouraged (alters renal clearance and perilymph dynamics).
Renal function must be verified before double or triple dose gadolinium. For single standard dose (0.1 mmol/kg), eGFR ≥ 30 mL/min/1.73 m² is the standard threshold. For double dose (0.2 mmol/kg), many centres require eGFR ≥ 45 mL/min/1.73 m². Verify local protocol and consultant nephrology guidance.
Macrocyclic GBCA is strongly preferred for all EH protocols, particularly for serial monitoring programmes. Linear GBCA agents have higher brain deposition rates with repeated use; macrocyclic agents (gadobutrol, gadoterate meglumine) are the preferred choice for patients who may undergo multiple EH MRI examinations over years of disease monitoring. This is consistent with the MRIninja-wide rule for serial monitoring programmes.
Prior cochlear implant: as per the CPA/inner ear master page, cochlear implant compatibility must be verified before any temporal bone MRI. The cochlear implant has no specific interaction with the EH protocol beyond its general compatibility requirement.
Metal near the ear: earrings, tragus piercings, any metallic items near the temporal bone must be removed. Their susceptibility artefacts extend precisely into the inner ear — the diagnostic target.
Prior surgical history: prior endolymphatic sac surgery, labyrinthectomy, or intratympanic injection history does not contraindicate EH MRI but may alter the expected anatomy and the expected EH pattern. Review the surgical history before interpreting.
3.2 Patient Positioning on the MRI System
Position: supine, head-first. Standard head coil (16–32 channel minimum). The positioning is identical to the CPA/inner ear protocol: isocentre at the level of the external auditory canal (tragus level), optimising B0 homogeneity at the temporal bones bilaterally.
This positioning requirement is non-negotiable for EH MRI. The 3D FLAIR and 3D T1 sequences at 4 hours post-Gd require the same B0 homogeneity at the inner ear level as the 3D CISS — placing the isocentre at the standard midface brain position degrades field homogeneity at the temporal bones and affects FLAIR null point accuracy.
Head symmetry: both temporal bones must be equidistant from the midline. Rotational asymmetry produces differential B0 homogeneity between the two inner ears, affecting the FLAIR TI null point differently on left vs right — making bilateral EH grading comparison unreliable.
Immobilisation: the 3D FLAIR at inner ear resolution (0.6–0.8 mm isotropic) over 5–7 minutes is highly motion-sensitive. The patient must be informed explicitly before starting the post-Gd sequences that absolute stillness is required. The patient has been waiting 4 hours at this point; fatigue-related restlessness should be anticipated and managed with comfortable head support.
4. Standard Protocol Design
The EH protocol is built around the 4-hour delayed gadolinium sequences as its primary diagnostic core, supported by structural sequences that provide the anatomical reference and exclude co-existing pathology.
4.1 Mandatory Core Sequences
| # | Sequence | Plane | Timing | Status |
|---|---|---|---|---|
| 1 | 3D CISS/DRIVE (0.5–0.7 mm isotropic) | Axial isotropic | Before or at standard post-Gd timing | Mandatory |
| 2 | T2 TSE | Axial | Before or after Gd | Mandatory |
| 3 | 3D FLAIR post-Gd (TI calibrated for 4h perilymph T1) | Axial isotropic | At 4 hours post-injection | Mandatory |
| 4 | 3D T1-weighted post-Gd | Axial isotropic | At 4 hours post-injection | Mandatory |
4.2 Conditional Sequences
| Sequence | Indication | Plane | Timing |
|---|---|---|---|
| Standard post-contrast T1-FS (5 min post-Gd) | First-ever EH protocol: vestibular schwannoma exclusion | Axial + coronal | 5 min post-injection |
| Pre-contrast 3D FLAIR | Reference for FLAIR comparison grading; some protocols require it | Axial isotropic | Before Gd |
| 3D T1 IT-Gd protocol | Intratympanic gadolinium variant; unilateral high-concentration | Axial isotropic | 30–90 min post-IT injection |
| DWI | First-ever EH protocol: exclude labyrinthine infarction | Axial | Before Gd |
| T1 axial (non-fat-suppressed) | Petrous apex assessment; cholesterol granuloma exclusion | Axial | Before Gd |
4.3 Rationale Summary Per Sequence
3D FLAIR Post-Gadolinium (4-hour delayed) — the Inverted Image Technique
This is the primary and most diagnostically specific EH sequence. Its mechanism requires detailed explanation because it is unique in all of clinical MRI.
The pharmacokinetic basis: intravenous gadolinium distributes into the perilymph via the blood-labyrinth barrier over 2–6 hours. At 4 hours post-injection, the perilymph gadolinium concentration produces a T1 of approximately 1200–2000 ms (from the baseline ~4000 ms). The endolymph, enclosed within the membranous labyrinth and separated from perilymph by the Reissner membrane and basilar membrane, does not accumulate gadolinium — its T1 remains at the baseline ~4000 ms.
The FLAIR mechanism: standard 3D FLAIR uses an inversion time (TI) designed to null the signal of CSF (T1 ≈ 4000 ms; null point TI ≈ 2700 ms, modified to TI ≈ 1700–1900 ms at clinical TR = 9000 ms). At 4 hours post-Gd, the endolymph still has T1 ~4000 ms and is nulled by the standard FLAIR TI — endolymph appears dark. The perilymph, with shortened T1 (~1200–2000 ms), is NOT nulled at the same TI — perilymph appears bright (partially or fully recovered Mz at the time of the readout excitation).
The image appearance: the inner ear on post-Gd FLAIR at 4 hours appears as bright perilymph surrounding dark endolymph. The total labyrinthine fluid space is bright (perilymph) with a dark void in the position of the endolymph.
In EH: the dark endolymph void is abnormally large relative to the total fluid-filled labyrinthine space. In severe EH (grade 2), the endolymph may occupy more than half the total vestibular fluid space — visible as a large dark region filling most of the vestibule.
TI calibration — the critical technical requirement: the standard brain FLAIR TI (1700–1900 ms at 3T) may also null the shortened-T1 perilymph if the perilymph T1 has been reduced to near the null point. To ensure that perilymph appears bright, the TI must be long enough that perilymph has recovered significantly but endolymph remains near zero. Published EH protocols at 3T use TI ≈ 2000–2400 ms. The exact value must be calibrated at each site based on the gadolinium dose and delay interval used, because these determine the actual perilymph T1 at the time of imaging.
A site that uses standard brain FLAIR TI (1700 ms) without recalibration will null both CSF and perilymph — producing a dark inner ear without EH differentiation. This is the most common implementation error in EH MRI programmes.
3D T1-weighted Post-Gadolinium (4-hour delayed) — the Direct Enhancement Technique
At 4 hours post-Gd, the 3D T1-weighted sequence (a short-TR GRE or TSE sequence providing T1 weighting) directly images the T1 shortening in perilymph. Perilymph appears T1-bright (gadolinium effect); endolymph appears T1-dark (no gadolinium). This is the “direct enhancement” view — the complementary perspective to the FLAIR inverted image.
The T1 sequence provides a more intuitive image than the FLAIR (bright perilymph = gadolinium present; dark endolymph = no gadolinium) and may be preferable for teaching and reporting. Some EH grading schemes use T1 measurements for semi-quantitative ratio calculations. The T1 and FLAIR sequences are complementary and both are standard in complete EH protocols.
3D CISS (structural reference)
As fully described in the CPA/inner ear master page. In the EH protocol, its role is: - Anatomical reference for interpreting the post-Gd FLAIR and T1 findings - Structural EH complications: labyrinthine membrane rupture; iatrogenic labyrinthine changes - Concurrent pathology: labyrinthine schwannoma; labyrinthine ossification; intralabyrinthine mass
The 3D CISS does not change with EH itself — EH is a fluid volume change, not a structural signal change visible on CISS. Its value is contextual.
T2 TSE
Standard brain T2 provides posterior fossa survey and excludes concurrent posterior fossa pathology (mass, MS plaque, vascular lesion) that might contribute to vestibulocochlear symptoms.
4.4 Sequence Matching and Cross-Sequence Consistency
The post-Gd 3D FLAIR and post-Gd 3D T1 must be acquired with identical geometry — same FOV, same voxel size, same slice positions and orientation. This is essential because EH grading compares the relative proportions of bright (perilymph) and dark (endolymph) volumes; any geometric inconsistency between the two sequences prevents valid comparison.
For serial monitoring, all parameters must be reproduced identically across examinations: same gadolinium agent; same dose (mmol/kg); same IV injection site (antecubital preferred for consistent bolus kinetics); same delay interval (4h ± 15 minutes); same 3D FLAIR TI; same voxel size. A change in any of these parameters between serial examinations introduces systematic differences in EH grade that are not attributable to disease change.
4.5 Fat Suppression
Post-Gd 3D FLAIR: no fat suppression. The standard brain 3D FLAIR protocol (no fat suppression) is used. Fat suppression would alter the magnetisation conditions at the labyrinth-petrous bone marrow interface and is not part of validated EH imaging protocols.
Post-Gd 3D T1: no fat suppression. The T1-bright signal of perilymph is the contrast mechanism; fat suppression is not used.
If a standard post-contrast T1-FS is acquired at 5 minutes (for vestibular schwannoma exclusion on first presentation): SPAIR or Dixon fat suppression at isocentre. This is a separate sequence from the EH-specific post-Gd sequences and follows standard CPA/inner ear post-contrast T1 protocol (see parent page).
Post-contrast STIR is absolutely contraindicated — as throughout all MRIninja protocols.
4.6 Slice Positioning — Complete Technical Reference
Why Precise Positioning Is Critical for EH MRI
The diagnostic content of EH MRI is entirely contained within the membranous labyrinth of both inner ears — structures measuring 3–8 mm in their largest dimension (cochlear duct width 0.5–1.0 mm; vestibule height 4–6 mm; saccule diameter 2–3 mm). These dimensions are comparable to or smaller than the voxel size of many MRI sequences. Sub-millimetre isotropic 3D acquisitions are required. Any positioning error that places the temporal bones off B0-isocentre degrades the FLAIR null point accuracy and risks non-diagnostic EH imaging.
Anatomical Landmarks
The anatomical targets for EH grading are:
Cochlea: 2.5 turns within the petrous bone. The cochlear duct (scala media / endolymphatic compartment) runs between the scala vestibuli and scala tympani. On post-Gd FLAIR: bright outer scalae (perilymph) surrounding dark cochlear duct (endolymph). EH grading uses the mid-cochlear turn cross-section where the three scalae are best resolved at 0.6–0.8 mm.
Vestibule: the ovoid chamber containing the utricle (upper) and saccule (lower). At 4h post-Gd: bright perilymph filling the vestibule with dark saccule and utricle endolymph visible within. Vestibular EH = the dark saccular/utricular endolymph space occupies an abnormally large proportion of total vestibular volume.
Saccule: the most sensitive and consistently enlarged structure in Ménière’s disease. Located in the inferior vestibule. Diameter normally 2–3 mm. In grade 2 EH, it may expand to fill the entire vestibule (6–8 mm), displacing the utricle.
Planning Sequence
- Standard three-plane localiser (identical to CPA/inner ear protocol)
- Plan 3D acquisitions to cover both temporal bones bilaterally, centred on the vestibule and cochlea
- Verify bilateral coverage on the axial localiser: both cochleae and vestibules fully within the 3D slab
Axial Planning
The primary acquisition plane for all EH sequences is axial isotropic. Planned parallel to the standard orbitomeatal line. The 3D slab must cover: - Inferior margin: below the round window (cochlear base) - Superior margin: above the superior semicircular canal - Lateral margin: includes both lateral aspects of the temporal bones - Medial margin: includes the CPA cisterns bilaterally (reference anatomy)
At 0.6–0.8 mm isotropic, the total slab for bilateral temporal bone coverage is approximately 3–4 cm craniocaudal × 12–15 cm lateral.
Phase encoding direction: A-P for axial acquisitions. Displaces any CSF pulsation artefacts anteroposteriorly, away from the inner ear structures.
Symmetry Verification
Before starting the post-Gd sequences, verify bilateral symmetry of the temporal bone positions on the axial localiser from the 3D CISS. If the head is rotated, the two inner ears will be at different distances from the isocentre and at different depths within the B0 field — producing asymmetric FLAIR null point performance. Reposition if asymmetry > 5 mm.
Serial Reproducibility
For serial EH monitoring examinations, the identical 3D slab position must be reproduced. Record: (a) the distance from the isocentre to the inferior 3D slab boundary; (b) the axial plane angle (degrees from orbitomeatal line). Many modern scanners allow saving the 3D CISS geometry as a reference for automatic re-positioning at follow-up.
Section 4.6 Dedicated Bibliography
Naganawa S, et al. Gadolinium-enhanced 3D-FLAIR detects endolymphatic hydrops: correlation with clinical findings in patients with Ménière’s disease. Eur Radiol. 2008;18(8):1689–1695. PMID: 18414866. DOI: 10.1007/s00330-008-0926-0. (Technical / Foundational) — Original description of 3D FLAIR post-Gd for EH; establishes bilateral coverage requirements, axial isotropic acquisition, and inner ear landmark methodology for EH positioning.
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) — 3T parameter optimisation; voxel size, TI calibration, and bilateral positioning requirements for inner ear EH FLAIR.
5. Optimisation Strategy
5.1 Artefact Reduction by Source
FLAIR TI miscalibration — the primary EH-specific failure mode: if the TI is set to the standard brain FLAIR value (1700–1900 ms at 3T), the gadolinium-containing perilymph is partially nulled alongside the endolymph and CSF. The inner ear appears uniformly dark — the EH pattern is lost. Conversely, if TI is too long (> 2500 ms at 3T), both compartments have substantially recovered and contrast between them is reduced. The site must calibrate TI specifically for the gadolinium dose and delay used. Published EH protocols use TI 2000–2400 ms at 3T for double-dose IV Gd at 4 hours. A test acquisition sweeping TI over this range in the first patients of a new EH programme is strongly recommended.
bSSFP banding artefact in 3D CISS (at 3T): as documented in the CPA/inner ear master page. The 3D FLAIR and T1 post-Gd sequences are TSE/GRE-based and do not have bSSFP banding. The banding risk applies only to the 3D CISS structural sequence. Management: adjust centre frequency before 3D CISS acquisition; use DRIVE (Philips) as banding-free alternative.
Motion during post-Gd 3D FLAIR: the patient has been waiting 4 hours before the diagnostic sequences begin. Fatigue-related restlessness during the 5–7 minute 3D FLAIR acquisition is common and degrades the sub-millimetre resolution required for EH grading. Mitigation: explicit verbal instruction immediately before starting the FLAIR; comfortable head immobilisation; consider shortest feasible acquisition time (CS-accelerated 3D FLAIR at 3T can reduce to 3–4 minutes).
Delay interval outside the 3.5–4.5 hour range: perilymph gadolinium concentration depends on the delay interval. At < 3.5 hours, perilymph T1 may not be sufficiently shortened for reliable FLAIR signal. At > 4.5 hours, gadolinium washout from perilymph (via CSF and renal clearance) may have begun, reducing perilymph T1 shortening. Document the actual delay interval and flag any deviation from the 3.5–4.5 hour range in the report.
Renal clearance variation: patients with mildly reduced renal function (eGFR 30–60 mL/min/1.73 m²) may show higher perilymph gadolinium concentration at 4 hours (reduced clearance → more gadolinium remaining in circulation → more in perilymph). This may actually improve EH signal quality. Conversely, patients with hyperdynamic renal function (athletes, well-hydrated young adults) may show faster gadolinium clearance and lower perilymph signal at 4 hours.
5.2 Protocol Efficiency and Throughput
The scanning time itself is modest (approximately 20–25 minutes at 3T for the full EH protocol). The 4-hour pharmacokinetic delay is the operational rate-limiting step. Throughput models for EH MRI programmes:
Model A — sequential same-day appointments: patient injected at 8:00; returns at 12:00; scanned 12:00–12:30. One patient per 4-hour injection slot. Maximum throughput: 2 patients per full working day on a dedicated EH scanner slot.
Model B — staggered injection clinic: 4–5 patients injected at 30-minute intervals from 9:00–11:00; scanned sequentially from 13:00–15:00. More efficient use of scanner time; requires a dedicated injection clinic with nursing staff.
Model C — intratympanic gadolinium (IT-Gd): eliminates the 4-hour wait. Patient injected intratympanically by ENT; imaged 30–90 minutes later. Higher inner ear concentration; unilateral only; invasive. For bilateral disease, bilateral IT-Gd is required. Most efficient for targeted unilateral assessment in a combined ENT-radiology setting.
5.3 Field Strength Considerations
3T is strongly recommended as the reference field strength for EH MRI: - Higher SNR enables 0.6–0.8 mm isotropic voxels for reliable cochlear duct and saccular resolution - The perilymph-endolymph signal differentiation on 3D FLAIR is more reliable at 3T (higher FLAIR SNR) - All published EH grading validation studies with surgical or histological correlation were performed at 3T [1, 4, 5]
1.5T is feasible but limited: - Reduced SNR forces larger voxels (0.8–1.0 mm or larger) → cochlear EH grade 1 may be undetectable - FLAIR performance at 1.5T for the inner ear requires re-calibration of TI (longer T1 values at 1.5T → different TI calibration from 3T) - Vestibular EH (saccular hydrops grade 2) is generally detectable at 1.5T; subtle grade 1 cochlear EH may be missed
SAR at 3T: the 3D FLAIR at inner ear resolution is a SAR-intensive sequence (TSE readout with long ETL, repeated RF pulses). At 3T, the SAR for a full-brain 3D FLAIR may require TR extension at maximum flip angle settings. For the targeted temporal bone 3D FLAIR (smaller slab), SAR is generally within acceptable limits.
6. Contrast Use Principles Specific to EH MRI
Gadolinium is not a supporting element of this protocol — it is the entire mechanism. Without gadolinium at the correct dose and delay, the examination cannot provide EH information.
6.1 Non-Contrast Standard Protocol — Sufficient For
The non-contrast components of the EH protocol (3D CISS, T2 TSE) provide only structural assessment of the labyrinth and exclusion of co-existing pathology. They provide no EH information. A non-contrast study cannot diagnose, grade, or exclude endolymphatic hydrops.
Non-contrast EH research approaches (high-resolution T2 at 7T, advanced diffusion techniques) have been described but are not clinically validated for EH diagnosis and grading.
6.2 Gadolinium — EH-Specific Dosing Framework
Single standard dose (0.1 mmol/kg IV): - Validated in the original Naganawa 2007–2008 series [1] - Provides adequate but sometimes marginal perilymph signal - May be insufficient for reliable grade 1 EH detection in the cochlea - Lowest gadolinium exposure; appropriate for first-line or monitoring in patients where cumulative exposure is a concern
Double dose (0.2 mmol/kg IV): - Most widely used in current clinical practice [5, 6] - Provides reliably bright perilymph signal on post-Gd FLAIR - Improved sensitivity for mild (grade 1) cochlear EH - Higher gadolinium exposure; requires eGFR ≥ 45 mL/min/1.73 m² at most centres - The dose recommended by most EH MRI consensus recommendations based on current evidence
Triple dose (0.3 mmol/kg IV): - Used in some research centres for maximal perilymph signal - No established additional diagnostic benefit over double dose in clinical series - Higher cumulative gadolinium exposure; generally not recommended for clinical monitoring programmes
Intratympanic gadolinium (IT-Gd): - Diluted gadolinium injected transtympanically into the middle ear - Diffuses through the round window into perilymph over 30–90 minutes - Achieves higher perilymph concentration than IV route - Unilateral by design (bilateral requires bilateral injection) - Imaging performed at 30–90 minutes post-IT injection (not 4 hours) - 3D T1 post-IT-Gd is the primary sequence (perilymph brighter than with IV Gd at 4h) - Invasive; requires ENT collaboration; not standard for bilateral assessment
Agent selection for serial monitoring: macrocyclic GBCA (gadobutrol, gadoterate meglumine) are mandatory for serial monitoring programmes due to lower CNS and tissue deposition compared with linear agents. This is a protocol-level requirement for all EH monitoring programmes, consistent with the MRIninja-wide rule for repeated GBCA administration.
6.3 Post-Contrast Acquisition Timing
The defining characteristic of this protocol: the diagnostic post-Gd sequences (3D FLAIR and 3D T1) are acquired at 4 hours (± 30 minutes) post-injection. This is the only timing that provides adequate differential T1 between perilymph and endolymph for EH grading.
If a standard vestibular schwannoma exclusion post-contrast T1-FS is required (first-ever EH examination): this is acquired at 5 minutes post-injection (standard equilibrium phase), before the patient leaves the department for the 4-hour wait. The imaging schedule: - T=0: gadolinium injection (double dose, macrocyclic) - T=5 minutes: standard post-contrast T1-FS (vestibular schwannoma exclusion) — optional, on first presentation only - Patient leaves; waits 4 hours - T=4 hours: patient returns; post-Gd 3D FLAIR + 3D T1
7. Reporting Essentials
7.1 The EH Grading System — Naganawa Scale
The universally used EH grading system was established by Naganawa et al. [1, 4] and is based on the proportion of the total labyrinthine fluid volume occupied by the dark endolymph space on post-Gd FLAIR or T1 images. Grading is performed separately for:
Cochlear EH (assessed on mid-cochlear turn cross-section): | Grade | Definition | Post-Gd FLAIR appearance | |—|—|—| | 0 — No EH | Endolymph ≤ 1/3 of total cochlear fluid space | Thin dark scala media; bright scalae tympani and vestibuli dominant | | 1 — Mild EH | Endolymph 1/3 – 1/2 of total cochlear fluid space | Dark scala media approaching the width of the adjacent scalae | | 2 — Significant EH | Endolymph > 1/2 of total cochlear fluid space | Dark cochlear duct dominates the cross-section; bright perilymph compressed |
Vestibular EH (assessed on axial section through the widest vestibular plane): | Grade | Definition | Post-Gd FLAIR appearance | |—|—|—| | 0 — No EH | Endolymph absent or minimal | Bright perilymph fills vestibule; no or tiny dark void | | 1 — Mild EH | Endolymph ≤ 1/2 of vestibular space | Dark saccular void occupying a minority of the bright vestibule | | 2 — Significant EH | Endolymph > 1/2 of vestibular space | Large dark saccular void occupying most of the vestibule; perilymph compressed to a rim |
Bilateral grading is mandatory: four separate grades are reported — right cochlear, right vestibular, left cochlear, left vestibular.
7.2 Interpretation Framework
The EH protocol is focused: the primary question is whether endolymphatic hydrops is present, in which compartment, at what grade, and bilaterally or unilaterally. Secondary questions include co-existing pathology (labyrinthine structural change, incidental mass) and perilymph signal adequacy (confirming that the protocol was technically adequate for grading).
Assessment axes: - Bilateral vs unilateral: bilateral EH (one or both ears positive) in 30–40% of clinical Ménière’s cases [3] - Cochlear-predominant vs vestibular-predominant vs combined: cochlear EH correlates with low-frequency hearing loss; vestibular EH (saccular) with episodic vertigo - Grade 0 bilateral: raises doubt about Ménière’s diagnosis; consider alternative aetiology - Grade 2 bilateral cochlear + vestibular: severe Ménière’s disease phenotype; bilateral disease management implications - Perilymph signal adequate (bright on FLAIR) vs absent (non-diagnostic delay or dose error): always assess and document
7.3 Mandatory Reporting Checklist
Technique documentation (mandatory in every EH MRI report): - [ ] Gadolinium agent (generic + brand name) - [ ] Dose: ___ mmol/kg = ___ mmol = ___ mL; route: IV / IT - [ ] Injection time: : - [ ] Imaging start time: : - [ ] Delay interval: ___ h ___ min (target: 4h ± 30 min; document if outside range) - [ ] 3D FLAIR TI used: ___ ms - [ ] Field strength; head coil; 3D voxel size
Right inner ear: - [ ] 3D CISS: labyrinthine structure — normal / abnormal (specify) - [ ] Post-Gd FLAIR: perilymph signal (bright) — adequate / inadequate (non-diagnostic) - [ ] Right cochlear EH grade: 0 / 1 / 2 - [ ] Right vestibular EH grade: 0 / 1 / 2 (saccular / utricular / combined)
Left inner ear: - [ ] Left cochlear EH grade: 0 / 1 / 2 - [ ] Left vestibular EH grade: 0 / 1 / 2
Overall: - [ ] Bilateral EH present: Yes / No - [ ] Pattern: cochlear-predominant / vestibular-predominant / combined / symmetric - [ ] Change from prior study (if serial): improvement / stable / progression (specify grades) - [ ] Co-existing pathology noted
7.4 Structured Reporting Template
Indication: Ménière’s disease — [clinical stage: probable/definite]; symptomatic ear [R/L/bilateral]; prior treatment history.
Technique: 3T head coil; [agent] [dose] mmol/kg IV/IT at [time]; 3D CISS 0.6 mm isotropic; 3D FLAIR (TI [ms]) post-Gd acquired at [time] (delay [X h Y min]); 3D T1 post-Gd.
Comparison: [prior EH MRI date; prior grades bilateral].
Findings — Inner Ear Assessment: Right cochlea: Grade /2. Right vestibule (saccule/utricle): Grade /2. Left cochlea: Grade /2. Left vestibule: Grade /2. 3D CISS: [normal structure bilaterally / abnormal — specify]. Post-Gd FLAIR perilymph signal: adequate bilaterally / [note if inadequate].
Impression: [Unilateral / Bilateral] endolymphatic hydrops. [Grade distribution summary]. [Change from prior if applicable]. [Clinical correlation recommended].
Limitations: delay interval [note if outside 3.5–4.5h]; [motion artefact if present]; [non-standard TI if used].
7.5 Incidental Findings — Clinical Decision Framework
Usually benign: mild asymmetric perilymph FLAIR signal between sides (typically reflects minor B0 asymmetry or minor flow differences; not clinically significant if bilateral EH grades are clear).
Requires communication: unexpected labyrinthine dark signal on CISS (possible labyrinthine ossification — relevant for cochlear implant candidacy if bilateral profound loss); unexpected IAC filling defect on CISS (vestibular schwannoma — report for standard CPA/inner ear work-up).
Urgent: unexpected brainstem or posterior fossa pathology on T2 (rare but possible as the sequences cover the posterior fossa).
8. MRI Technologist Pearls
8.1 Sequence Order Logic
The 4-hour pharmacokinetic constraint drives the sequence order. The post-Gd 3D FLAIR and T1 must be acquired at exactly the correct delay. All structural sequences must be completed before the patient leaves after injection (if possible), so that when the patient returns at 4 hours, scanner time is dedicated exclusively to the diagnostic EH sequences.
Full EH protocol sequence order:
| Step | Sequence | When | Notes |
|---|---|---|---|
| 1 | 3D CISS (0.6 mm isotropic) | Before or after injection (no Gd effect on CISS) | Structural reference; complete before 4h window |
| 2 | T2 TSE axial | Before injection | Standard structural survey |
| 3 | T1 axial (no FS) | Before injection | Petrous apex; baseline |
| 4 | DWI (if first exam) | Before injection | Exclude infarction/epidermoid |
| 5 | Gadolinium injection | T=0 | Record exact time; start 4h timer |
| 6 | Standard post-contrast T1-FS (optional) | T+5 min | First presentation only: schwannoma exclusion |
| 7 | Patient leaves department | T+10 min – T+3h 50min | Wait period |
| 8 | Patient returns | T+3h 50min | Re-verify position; re-shim |
| 9 | Post-Gd 3D FLAIR (calibrated TI) | T+4h ± 30 min | Primary EH sequence; document start time |
| 10 | Post-Gd 3D T1 | Immediately after FLAIR | Direct enhancement sequence |
8.2 Positioning Tricks
Re-shim when the patient returns at 4 hours: the patient leaves and returns to the scanner for a different session. Body position between sessions will differ marginally. Always re-shim the B0 field at the temporal bone level after the patient is repositioned at 4 hours. A poorly shimmed 3D FLAIR at 4 hours is non-diagnostic regardless of the correct delay.
Isocentre verification: at 4 hours, re-confirm that the isocentre is at the tragus level. If a brain survey was performed earlier with the isocentre at midface, the couch position must be adjusted before the post-Gd FLAIR.
Communication with the patient before the FLAIR: immediately before starting the 3D FLAIR, say explicitly: “This next sequence is 6 minutes. Please do not move at all during this time — it is the most important part of the examination. If you need to swallow or adjust, do it now, before I start.” Patient understanding and active co-operation at this specific moment substantially reduces motion artefact in a fatigued patient.
8.3 Fast Salvage Protocol
If the patient cannot complete the full post-Gd phase within the 4-hour window (delay exceeded, patient unable to remain):
| Priority | Sequence | Time | What it covers |
|---|---|---|---|
| 1 | Post-Gd 3D FLAIR | 5–6 min | Primary EH grading sequence — only one needed if time is critical |
| 2 | Post-Gd 3D T1 | 4–5 min | Complementary direct enhancement |
If only one sequence can be acquired at the 4-hour window, the post-Gd 3D FLAIR is the priority. It alone provides adequate EH grading in the majority of cases.
8.4 Common Avoidable Errors
| Error | Consequence | Prevention |
|---|---|---|
| FLAIR TI not recalibrated for post-Gd perilymph | Perilymph nulled along with CSF; inner ear uniformly dark; non-diagnostic EH examination | Use site-validated TI (2000–2400 ms at 3T for double-dose IV Gd at 4h); never use standard brain FLAIR TI (1700–1900 ms) for EH imaging |
| Injection time not recorded precisely | Delay interval uncertain; FLAIR acquired outside optimal window without knowing it | Record injection time (HH:MM) in the scan request immediately at the moment of injection; not estimated retrospectively |
| Patient imaged too early (< 3.5 hours) | Perilymph T1 not sufficiently shortened; FLAIR signal inadequate; EH under-graded or missed | Enforce the 4-hour delay; do not accept the patient early; if patient arrives early, wait |
| Unilateral coverage (symptomatic ear only) | Clinically silent contralateral EH missed in 30–40% of cases; ablative therapy decisions based on incomplete bilateral assessment | Always image bilaterally; both inner ears must be fully within the 3D FLAIR and T1 coverage |
| Linear GBCA used for serial monitoring programme | Cumulative gadolinium deposition over multiple examinations | Use macrocyclic GBCA for all EH protocols; this is mandatory per MRIninja protocol standard for serial monitoring |
| Post-Gd 3D FLAIR acquired with standard brain FLAIR protocol fat-suppressed | Fat suppression alters the FLAIR magnetisation at the temporal bone region | No fat suppression on post-Gd 3D FLAIR for EH |
| Patient not re-positioned and re-shimmed at 4 hours | B0 inhomogeneity at temporal bones from different head position; FLAIR null point inaccurate | Always re-shim at 4 hours after patient is repositioned |
9. Quality Control Checklist
10. Advanced Technical Parameters
10.1 Post-Gadolinium 3D FLAIR — Inner Ear Specific
The TI Calibration Problem — Physical Explanation
The FLAIR null point for a tissue with T1 is:
TI_null = T1 × ln(2) × [1 − exp(−TR/T1)]^(−1)
For long TR (TR >> T1): TI_null ≈ T1 × ln(2) ≈ 0.693 × T1
At 3T: - Normal CSF (T1 = 4000 ms): TI_null ≈ 2770 ms; at clinical TR = 9000 ms: TI ≈ 1900 ms - Post-Gd perilymph (T1 = 1500 ms): TI_null ≈ 1040 ms; at TR = 9000 ms: TI ≈ 1040 ms × correction ≈ 1000 ms
To null endolymph (T1 ≈ 4000 ms) while NOT nulling perilymph (T1 ≈ 1500 ms at double dose, 4h): - Use TI ≈ 1900 ms: nulls CSF/endolymph ✓; perilymph (T1=1500 ms) is already recovering → perilymph Mz > 0 → perilymph appears bright ✓
However, perilymph T1 varies between patients (gadolinium dose, renal function, blood-labyrinth barrier permeability). A patient with higher perilymph Gd (shorter T1 ~1000 ms) needs a longer TI to avoid nulling perilymph. A patient with lower perilymph Gd (T1 ~2000 ms) needs a shorter TI.
Practical calibration: use TI 2000–2200 ms at 3T for double-dose IV Gd at 4h as the centre of the calibration range. Optimise at each institution by comparing post-Gd inner ear signal in the first 5–10 patients.
Key Parameters
| Parameter | 1.5T | 3T | Rationale |
|---|---|---|---|
| Sequence type | 3D TSE FLAIR | 3D TSE FLAIR | Standard FLAIR readout; vendor product |
| TI | 2200–2600 ms | 2000–2400 ms | Calibrated for post-Gd perilymph T1 (longer at 1.5T: longer T1 values at lower field) |
| TR | 9000–12000 ms | 7000–9000 ms | Long TR for full longitudinal recovery |
| ETL | 100–200 | 100–200 | 3D FLAIR standard long ETL |
| Voxel size | 0.8–1.0 mm isotropic | 0.6–0.8 mm isotropic | Cochlear duct resolution requirement |
| Fat suppression | None | None | Not used |
| Parallel imaging | R=2 (+ CS if available) | R=2–3 (CS preferred) | CS reduces acquisition time to 3–4 min |
| Acquisition time | 6–8 min | 4–6 min (standard R=2); 3–4 min (CS) |
Vendor implementations: - Siemens: product 3D FLAIR (SPACE); modify TI from the standard 1800 ms to the calibrated EH value - GE: 3D CUBE FLAIR; TI modification in the IR preparation tab - Philips: product 3D FLAIR (VISTA or IR-TSE); TI is directly accessible
10.2 Post-Gadolinium 3D T1 — Direct Enhancement Technique
| Parameter | 1.5T | 3T | Rationale |
|---|---|---|---|
| Sequence type | 3D GRE (MPRAGE-variant) or 3D T1 TSE | 3D GRE or 3D T1 TSE | T1-weighting; gadolinium T1 shortening |
| TR | 5–15 ms (GRE) or 500–800 ms (TSE) | 3–8 ms (GRE) or 400–600 ms (TSE) | Short TR for T1 weighting |
| TE | Shortest available | Shortest available | Minimise T2* effects |
| Voxel size | 0.8–1.0 mm isotropic | 0.6–0.8 mm isotropic | Same as FLAIR for direct comparison |
| Fat suppression | None | None | Not used |
Section 10 Dedicated Bibliography
Naganawa S, et al. Gadolinium-enhanced 3D-FLAIR detects endolymphatic hydrops: correlation with clinical findings in patients with Ménière’s disease. Eur Radiol. 2008;18(8):1689–1695. PMID: 18414866. DOI: 10.1007/s00330-008-0926-0. (Technical / Foundational) Original 3D FLAIR technique; TI selection and pharmacokinetic basis; primary technical reference for all EH FLAIR protocols.
Naganawa S, et al. Semi-quantification of endolymphatic hydrops using MR cisternography after intravascular gadolinium injection. Magn Reson Med Sci. 2008;7(3):117–123. PMID: 18997438. DOI: 10.2463/mrms.7.117. (Technical / Foundational) Grading system validation; direct T1 technique for EH semi-quantification; reference for grade 0–2 definitions.
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) 3T parameter optimisation: TI calibration, voxel size, and field-strength-specific adjustments.
11. Evidence Gaps and Ongoing Debate
Optimal gadolinium dose: single (0.1), double (0.2), and triple (0.3 mmol/kg) IV doses have all been published. No RCT compares EH grading accuracy vs dose. Double dose has the most published data and appears to be the current clinical consensus but is not based on a formal head-to-head trial.
IV vs IT gadolinium — diagnostic equivalence: IT-Gd achieves higher perilymph concentration than IV and eliminates the 4-hour delay. No prospective comparative trial has established diagnostic equivalence or superiority between IV-4h and IT-Gd for EH grading against a clinical or histological reference standard.
TI calibration standardisation: published EH protocols use different TI values (1800–2400 ms at 3T). No multi-site calibration study has established a consensus TI for standardised gadolinium dose and delay. This makes inter-institution EH grade comparisons in multicentre studies methodologically problematic.
EH grade reproducibility: inter-reader agreement for Naganawa grades is moderate (κ 0.5–0.7 in published series). Whether automated AI-based grading (volumetric endolymph/perilymph ratio from segmented 3D FLAIR) would improve reproducibility has been studied in limited research series but is not clinically validated.
Clinical correlate of EH grade: EH grade does not correlate tightly with clinical disease severity, pure-tone audiogram, or attack frequency. Grade 2 vestibular EH may be found in mild Ménière’s; grade 1 cochlear EH may accompany severe low-frequency hearing loss. The EH grade is most valuable as a presence/absence and bilateral distribution marker rather than as a direct severity metric.
Optimal follow-up interval for serial monitoring: no published trial has established the optimal EH MRI interval after treatment (6 months? 12 months? After each treatment cycle?). Current practice is centre-dependent.
12. Evidence-Based References
A. Guidelines / Consensus / Society Recommendations
B. Systematic Reviews / Meta-analyses
C. Important Prospective / Original Studies
D. Technical MRI Papers
End of document — MRI Endolymphatic Hydrops Generic Standard Protocol — MRIninja v1.0 — May 2026 Related pages: MRI CPA and Inner Ear Generic Standard Protocol · MRI Middle Ear and Petrous Bone Generic Standard Protocol Future child pages: Ménière’s disease definite/probable staging; secondary EH (autoimmune, post-viral, post-traumatic); IT-Gd protocol; bilateral EH treatment monitoring; EH MRI in superior canal dehiscence
Child Protocols
Clinical pages derived from this master protocol. These pages document what changes for specific indications.
No child protocols have been published yet.
Related Protocols
Recent PubMed search for this protocol