MRIninja Knowledge Base | Master / General Protocol Page Related pages: Soft Tissues Neck MRI · Brachial Plexus MRI Version 1.0 — May 2026

1. Executive Summary

Parotid gland MRI is the definitive preoperative imaging investigation for parotid masses and is the modality of choice for parotid pathology that requires soft tissue characterisation beyond the capability of ultrasound. The parotid gland — the largest of the three major salivary glands — is uniquely accessible to ultrasound (superficial lobe), but its deep lobe, the parapharyngeal extension, the facial nerve course through the gland, the retromandibular vein, and the parotid duct are all either inaccessible or inadequately characterised by ultrasound alone. MRI provides the definitive preoperative assessment of the anatomical extent of parotid pathology, the relationship to the facial nerve plane, deep lobe involvement, and any associated regional lymphadenopathy or perineural spread — information that directly determines the surgical approach (superficial parotidectomy, total parotidectomy, nerve-sparing or nerve-sacrificing).

CT provides superior bone detail and is faster; ultrasound provides real-time assessment with guided biopsy capability. Neither modality can characterise the full extent of deep lobe tumours, map perineural spread along CN VII, or discriminate between histological subtypes of parotid tumours with the precision of MRI. The ACR Appropriateness Criteria [1] designate MRI as the imaging modality of choice for parotid mass characterisation when surgical management is planned.

The generic parotid MRI protocol is designed for bilateral parotid gland assessment — providing complete characterisation of both glands simultaneously, with systematic evaluation of both the superficial and deep lobes, the intraparotid facial nerve plane, adjacent lymph nodes, the parapharyngeal space, and the mandible. The dedicated target coil centred on one or both parotids provides the SNR needed for the 0.5–0.7 mm in-plane resolution required to characterise the internal architecture of parotid masses.

1.1 Core Strengths

Deep lobe and parapharyngeal space assessment: the deep lobe of the parotid gland lies medial to the stylomandibular tunnel, separated from the parapharyngeal space by the stylomandibular ligament. Deep lobe tumours — which constitute approximately 15–25% of parotid tumours — are not accessible to ultrasound and require MRI for accurate delineation. On axial and coronal T2 and T1 sequences, the deep lobe and any parapharyngeal extension are fully characterised.

Facial nerve plane identification: while MRI cannot directly visualise the individual branches of CN VII within the parotid (< 1 mm, below routine MRI resolution), the retromandibular vein — which runs in the same tissue plane as the main trunk of CN VII — serves as a reliable surrogate landmark for the facial nerve plane. The MRI demonstrates: (a) the position of the mass relative to the retromandibular vein (superficial vs deep lobe); (b) any displacement or invasion of the retromandibular vein; (c) any T2 signal change along the expected facial nerve course suggesting perineural spread; (d) stylomastoid foramen and mastoid segment of CN VII for proximal extension assessment.

Perineural spread along CN VII: parotid malignancy can spread proximally along CN VII — through the stylomastoid foramen to the mastoid segment, through the facial canal to the geniculate ganglion, and intracranially to the CPA cistern. MRI with post-contrast T1 and dedicated coronal and axial sequences through the facial canal is the only modality that detects this proximal perineural spread, which fundamentally changes the surgical approach and prognosis.

Tissue characterisation for differential diagnosis: MRI signal characteristics enable pre-operative differential diagnosis that ultrasound and CT cannot provide: - Pleomorphic adenoma (benign mixed tumour): T2-very bright (myxoid stroma), T1-dark, well-defined lobulated margin, no restricted diffusion, delayed persistent enhancement - Warthin tumour: T2-intermediate with internal T2-dark foci (lymphocytes), bilateral in 20%, T1-intermediate, moderate enhancement, restricted diffusion (low ADC due to cellular content) - Mucoepidermoid carcinoma: T2-variable, ill-defined margin, may restrict diffusion; overlaps with benign lesions - Adenoid cystic carcinoma: T2-intermediate, perineural spread, ill-defined margin; often T2-dark from fibrous content

Bilateral gland assessment: both parotid glands are imaged simultaneously. Bilaterality (bilateral Warthin tumours, bilateral inflammatory enlargement in Sjögren, bilateral parotitis) is immediately apparent without additional examination time.

Parotid duct and accessory parotid assessment: the Stensen duct (parotid duct, approximately 4–5 cm long) runs anteriorly from the parotid gland, over the masseter muscle, to enter the oral cavity at the upper second molar. Accessory parotid tissue along the duct course is visible on MRI and must be assessed as a potential site for tumour recurrence or primary accessory gland lesions.

1.2 Intrinsic Limitations of the Generic Protocol

CN VII direct visualisation: the main trunk and branches of the facial nerve within the parotid gland (1–2 mm diameter) are below the spatial resolution of standard clinical MRI. The facial nerve plane is inferred from the retromandibular vein position, not directly visualised. In cases where the retromandibular vein is displaced or absent, the nerve plane inference is unreliable.

Histological specificity: MRI can narrow the differential diagnosis of parotid masses and assign a level of malignant suspicion, but cannot provide histological diagnosis. Fine-needle aspiration cytology (FNAC) or core biopsy, guided by ultrasound, remains necessary for definitive preoperative diagnosis in most cases. The MRI differential diagnosis guides clinical management but does not replace pathological confirmation.

Dental metalwork: as with all parotid-adjacent examinations, dental amalgam restorations in the upper and lower jaw produce susceptibility artefacts that can extend into the posterior oral cavity and parotid region. At 3T, these artefacts are substantially larger than at 1.5T and may degrade assessment of the deep lobe medial margin and the parapharyngeal space in patients with extensive dental metalwork.

Dynamic sialography: MRI cannot provide functional assessment of ductal drainage (which requires MR sialography — a dedicated contrast-based or lemon juice-stimulated ductal filling technique). For suspected parotid duct strictures, stones (though MR sialography or CT sialography is needed for calculus characterisation), and Sjögren’s ductal assessment, dedicated MR sialography is a supplement to the generic protocol.

When dedicated child protocols are required: parotid malignancy staging with perineural spread mapping (requires dedicated facial canal and skull base sequences); Sjögren syndrome staging (requires specific ductal and parenchymal scoring); MR sialography for duct assessment; parotid lymphoma (full neck protocol required); parotid involvement in IgG4-related disease; paediatric parotid (age-specific protocol); fistula-in-ano parotid abscess with trismus.

2. Main Clinical Indications

2.1 Standard Indications

Parotid mass characterisation is the dominant indication. When ultrasound identifies a parotid mass and surgical management is planned, MRI is required for: deep lobe extent; retromandibular vein relationship (facial nerve plane); parapharyngeal extension; internal architecture (T2 signal, internal complexity); and perineural spread screening. The generic protocol is adequate for initial characterisation of most parotid masses.

Bilateral parotid enlargement — from Sjögren syndrome, sarcoidosis, HIV-associated parotitis, metabolic/nutritional parotitis (alcoholism, bulimia), and viral parotitis — is assessed by MRI when ultrasound is inconclusive or systemic diagnosis requires comprehensive assessment. MRI provides bilateral parenchymal signal assessment (diffuse T2 signal change in Sjögren), lymphoepithelial cysts (T2-bright bilateral intraparotid cysts in HIV), and lymphomatous transformation in Sjögren (focal T2-dark masses in a background of Sjögren changes).

Parotitis assessment — acute and recurrent — requires MRI when ultrasound cannot adequately assess abscess vs phlegmon or when deep space involvement is suspected. Post-contrast T1 delineates the rim-enhancing abscess within the gland or the diffuse enhancement of phlegmonous parotitis. Ludwig’s angina extending from the submandibular space to the parotid requires full neck protocol.

Post-surgical assessment after parotidectomy: MRI detects: recurrent pleomorphic adenoma (can occur 15–30 years after surgery at the implantation site); recurrent malignancy; surgical bed changes; scar vs viable tumour. Post-surgical fat grafts and clips produce susceptibility and T1-bright signal changes that must be distinguished from recurrence.

Incidental parotid lesion on CT or ultrasound: when an incidental lesion is found in the parotid on another examination, MRI provides characterisation that may prevent unnecessary biopsy (classic pleomorphic adenoma or Warthin tumour appearance) or identify features requiring surgical management.

2.2 Urgent Red Flags Requiring Expedited or Emergency Imaging

The parotid gland is not a life-threatening organ. However:

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

Dental metalwork documentation: dental amalgam, crowns, bridges, and implants in the posterior maxilla and mandible produce susceptibility artefacts that can extend into the deep parotid and parapharyngeal space. At 3T, these artefacts may be 2–5 cm in radius, obscuring the deep lobe medial margin and the stylomandibular tunnel. Document all dental metalwork. For patients with extensive posterior dental metalwork affecting the target area, consider 1.5T.

Earrings and periauricular jewellery: earrings, tragus piercings, and other periauricular jewellery are immediately adjacent to the parotid gland. These must be removed. Metallic earring backs can produce susceptibility artefacts directly over the superficial lobe.

Hearing aids: remove before the examination (see general preparation page). Behind-the-ear hearing aids may contain metallic components positioned directly adjacent to the parotid.

Prior parotidectomy: document the surgical history — specifically: which side, extent (superficial vs total parotidectomy), type of reconstruction (fat graft, free flap), CN VII identification/sacrificed. The post-surgical anatomy will be completely altered on the operated side. Fat grafts appear T1-bright and STIR-suppressed, simulating normal fat but located in the surgical bed. Surgical clips and haemostatic agents produce T2* signal loss.

Sialendoscopy clips or stents: some patients presenting for MRI sialography will have had prior sialendoscopy. Metallic clips or stents in the duct system produce artefacts; document.

Fasting for MR sialography (conditional): if MR sialography is planned, no specific fasting is required; lemon juice or citric acid stimulation is administered immediately before the MRCP-equivalent heavily T2-weighted duct sequence.

3.2 Patient Positioning on the MRI System

Position: supine, head-first, head in a neutral position — neither rotated, flexed, nor extended. Head rotation is the most common positioning error for bilateral parotid MRI: even 5° of head rotation produces asymmetric coil coupling between the two parotids and asymmetric fat suppression quality, making bilateral comparison unreliable.

Coil selection: the choice of coil determines the spatial resolution achievable for parotid assessment:

• Head coil (16–32 channel) provides bilateral coverage at a consistent isocentric position. This is the standard coil for most departments and provides adequate SNR for 0.5–0.7 mm in-plane resolution. The head coil FOV (24–28 cm) provides complete bilateral parotid coverage.

• Head-and-neck combination coil: provides improved posterior coverage, particularly useful when the tail of the parotid and the upper neck lymph nodes require simultaneous assessment. The standard configuration for departments without a dedicated head coil.

• Surface coil (bilateral flexible) placed directly over both parotids: provides the highest SNR for targeted bilateral assessment but is only appropriate when both glands are the exclusive target and no adjacent neck assessment is required.

• Single-sided flexible surface coil: for unilateral parotid lesion with excellent clinical localisation, a dedicated surface coil on the affected side provides maximum SNR — useful for very high resolution (0.3–0.4 mm) targeted assessment of a specific lesion, but sacrifices contralateral comparison.

Centring: isocentre at the level of the parotid glands — approximately the level of the external auditory canals. Verify on the three-plane localiser that both parotid glands are symmetrically positioned within the FOV — the mastoid processes should appear symmetric bilaterally on the coronal localiser; the mandibular angles should be at the same height on the axial localiser.

Head symmetry verification: before starting the diagnostic sequences, verify on the axial localiser that: (a) the mandibular condyles are at the same height bilaterally; (b) the distance from the midline to each mandibular ramus is equal; (c) the nasal septum is midline. Asymmetry indicates head rotation — correct before starting.

Immobilisation: a head holder (standard head coil cradle) maintains head position. Foam pads lateral to the head reduce involuntary movement. For patients with tremor or restlessness, firm foam pads may be placed bilaterally adjacent to the temporal bones.

Common positioning errors: - Head rotated → asymmetric coil coupling → asymmetric signal and fat suppression → bilateral comparison unreliable - Chin elevated → parotid tails displaced inferiorly → inferior parotid tail and angle of mandible nodes out of coverage - Chin depressed → submandibular glands overlap the deep parotid on axial sections

4. Standard Protocol Design

4.1 Mandatory Core Sequences

4.2 Conditional Sequences

4.3 Rationale Summary Per Sequence

Axial T2 non-fat-suppressed provides the fundamental anatomical reference for both parotid glands. On non-fat-suppressed T2: the parotid gland itself appears as a moderately T2-bright structure against the adjacent subcutaneous fat; the deep lobe is clearly demarcated by the stylomandibular tunnel; the retromandibular vein is visible as a dark (flow void) tubular structure within the gland, marking the surgical facial nerve plane. This non-fat-suppressed T2 is the sequence on which the lesion’s position relative to the retromandibular vein — and therefore its superficial vs deep lobe location — is most accurately determined. Fat suppression removes the T2-bright fat that creates this contrast; therefore, the non-fat-suppressed T2 is not optional — it is the anatomical complement to the fat-suppressed T2.

The parotid gland signal on non-fat-suppressed T2 is distinctly brighter than the adjacent masseter muscle (parotid has a relatively fatty parenchyma, particularly in adults, contributing to its T2 brightness), which provides the background parenchymal reference for identifying T2-abnormal foci.

Axial T2 fat-suppressed (SPAIR or STIR) reveals pathological signal within the gland parenchyma and adjacent tissues. On fat-suppressed T2: normal parotid parenchyma appears intermediate signal; lesions with T2-prolonged signal (cysts, mucinous/myxoid tumours, oedema) appear bright; T2-dark lesions (densely cellular tumours, fibrosis, haemosiderin) appear dark. This sequence is critical for: (a) identifying the T2 signal of the mass (the primary tissue characterisation step); (b) perineural tissue along the facial nerve course (T2-bright = oedema or tumour); (c) lymph node signal in the parotid lymph nodes (both intraparotid nodes and periparotid level II nodes); (d) parapharyngeal fat signal (asymmetric obliteration of parapharyngeal fat suggests deep lobe extension).

Coronal T2 fat-suppressed provides the longitudinal view of the bilateral parotid glands, the periparotid lymph nodes (levels I and II), and the parotid tails. The coronal plane is the most informative single plane for assessing: - Inferior extension of the gland and tail (accessory parotid nodes at angle of mandible) - Bilateral comparison of gland size and signal (Sjögren: bilateral increased signal; unilateral asymmetry: tumour or inflammation) - The relationship of the superior parotid to the external auditory canal and the mastoid process

Axial T1 non-fat-suppressed provides the T1 map for lesion characterisation. The parotid gland appears T1-bright in adults due to fat lobules within the gland parenchyma — more so than other salivary glands. T1-dark lesions against this T1-bright background are therefore conspicuous. Lesions that are T1-bright (indicating fat, haemorrhage, or proteinaceous content) are identified on this sequence. The T1 also provides: bone marrow signal in the mandible and the temporal bone; the styloid process and stylomandibular ligament as dark structures; vascular signal (internal carotid and jugular vein in the adjacent carotid space).

DWI and ADC for parotid assessment: DWI has proven particularly useful in parotid imaging for one specific reason — Warthin tumours (the second most common benign parotid tumour) show markedly restricted diffusion (very low ADC, approximately 0.7–1.0 × 10⁻³ mm²/s) due to their densely packed lymphocytic stroma. Pleomorphic adenomas have intermediate-to-high ADC (1.5–2.5 × 10⁻³ mm²/s) due to their loose myxoid matrix. Malignant parotid tumours generally show restricted diffusion (low ADC). This ADC discriminator — while not absolute — significantly improves the pre-operative differential diagnosis [4, 5]:

The overlap between Warthin tumour ADC and malignant lesions ADC is the primary limitation of DWI in parotid imaging. However, the combination of T2 signal + morphology + ADC provides a meaningful pre-operative risk stratification.

Post-contrast T1 fat-suppressed is the definitive enhancement characterisation sequence for parotid masses. Enhancement pattern: - Pleomorphic adenoma: progressive delayed enhancement (time-intensity curve type I — persistent); early post-contrast (1–2 min) shows less enhancement than late post-contrast (5–8 min) - Warthin tumour: rapid early enhancement with plateau or mild washout (type II) - Malignant tumours: variable; often early enhancement with washout - Abscess: rim enhancement; non-enhancing central zone

Perineural spread along CN VII is detected as asymmetric enhancement along the expected nerve course, within the stylomastoid foramen, the facial canal, or at the geniculate ganglion — visible only on post-contrast T1 sequences.

4.4 Sequence Matching and Cross-Sequence Consistency

The axial T2 non-FS, T2-FS, T1, and post-contrast T1 must use the same axial slice geometry — identical level positions, thickness, and FOV — to enable pixel-by-pixel comparison across sequences. For parotid lesion characterisation, the radiologist compares the lesion’s T2 signal, T1 signal, enhancement pattern, and ADC at the same anatomical level. Any geometric mismatch makes this multi-parametric analysis unreliable.

The coronal sequences (T2-FS, post-contrast T1-FS) should also use matching geometry for bilateral comparison.

For serial follow-up, document: coil type, FOV, slice thickness, and the reference anatomical level (e.g., “axial slices perpendicular to the tragal line, with the superior slice at the level of the external auditory canal meatus”). This reference enables exact geometric reproduction at follow-up.

4.5 Fat Suppression in Parotid MRI

SPAIR vs STIR for parotid T2-FS: the parotid glands are at or near isocentre when a head or head-and-neck coil is used and the patient is correctly positioned. At isocentre, B0 homogeneity is adequate for spectral fat suppression (SPAIR). SPAIR provides higher SNR than STIR for equivalent acquisition time and is the preferred fat suppression technique for parotid T2 sequences at isocentre. STIR is reserved for patients with significant B0 inhomogeneity from dental metalwork or when SPAIR fails.

At 3T, SPAIR remains preferred at isocentre but Dixon provides better B0-independent suppression when dental metalwork is present. The choice between SPAIR and Dixon at 3T is centre-dependent.

T1 fat-suppressed post-contrast: Dixon is preferred at 3T for post-contrast T1 to ensure uniform fat suppression near dental metalwork and in the deep parotid and parapharyngeal regions. SPAIR is acceptable when dental metalwork is minimal.

Non-fat-suppressed T2 is mandatory: as with the neck and brachial plexus protocols, the non-fat-suppressed T2 is required as a complement to the fat-suppressed T2. The non-fat-suppressed T2 shows the retromandibular vein, deep lobe delineation, and parapharyngeal fat anatomy that are obscured on fat-suppressed sequences.

4.6 Slice Positioning — Complete Technical Reference

Why Parotid Slice Positioning Is Critical

The parotid gland occupies the retromandibular and preauricular space, extending from the level of the external auditory canal (EAC) superiorly to the angle of the mandible inferiorly. The standard anatomical axes for parotid imaging are:

• Axial: perpendicular to the cervical spine / parallel to the orbital line — the standard head axial plane
• Coronal: perpendicular to the standard axial, parallel to the mandibular rami
• Optional sagittal oblique: parallel to the long axis of the Stensen duct

Anatomical Landmarks

External auditory canal (EAC): the superior limit of the parotid gland is at the EAC level. The superior coverage must include the EAC to assess EAC involvement by parotid tumours, which changes staging and surgical approach.

Mastoid process and stylomastoid foramen: the facial nerve exits the skull at the stylomastoid foramen on the anterior surface of the mastoid process. For perineural spread assessment, the facial canal within the temporal bone must be included in the superior coverage.

Retromandibular vein: within the parotid, just deep to the gland substance on the standard axial plane, the retromandibular vein is visible as a rounded dark (flow void) structure at the level of the mandibular ramus. This is the surgical facial nerve landmark.

Stylomandibular tunnel: the space between the styloid process (medial) and the mandibular ramus (lateral), through which the deep lobe extends into the parapharyngeal space. Visible on axial sequences at the level inferior to the mandibular condyle.

Masseter muscle: anterior to the parotid on axial images; the Stensen duct courses over the masseter superficially.

Angle of the mandible: the inferior limit of the parotid tail. The inferior coverage must include the angle of the mandible to assess the parotid tail and the periparotid lymph nodes at this level.

Planning Sequence

1. Three-plane localiser (standard head MRI scan plane)
2. From the sagittal and coronal localisers: confirm that the EAC is visible superiorly and the mandibular angle is visible inferiorly
3. Plan all axial sequences from a common reference: slices perpendicular to the cervical spine at the upper neck level, with the superior slice at the EAC level

Axial Planning

Reference: the sagittal localiser. Draw the axial slice prescription perpendicular to the long axis of the cervical spine at the C2–C3 level. This provides sections parallel to the standard head axial plane.

Coverage: from the level of the EAC (superiorly) to the angle of the mandible (inferiorly). For perineural spread assessment, extend the coverage superiorly to include the mastoid process and the stylomastoid foramen.

Phase encoding direction: A-P (anterior-posterior) for axial parotid sequences. This displaces any swallowing or vascular artefacts anteroposteriorly — away from the lateral parotid gland and the mandibular ramus.

Slice thickness: 3–4 mm for standard assessment; 2–3 mm for targeted high-resolution assessment of a specific lesion.

FOV: 200–240 mm is adequate for bilateral parotid assessment. A larger FOV (260–280 mm) is required when the neck nodes (levels I–II) must be included in the same acquisition.

Coronal Planning

Reference: the axial localiser. Plan the coronal prescription parallel to the posterior ramus of the mandible (visible on the axial localiser at the level of the mandibular condyle). This provides sections that show the full length of the bilateral parotid glands from superiorly to the parotid tail.

Phase encoding direction: R-L (right-to-left) for coronal parotid sequences. This displaces motion artefacts laterally.

Coverage: from the masseter muscle anteriorly to the prevertebral muscles posteriorly (approximately 6–8 cm in AP extent). This ensures that the full mediolateral extent of the parotid and parapharyngeal space is covered.

Perineural Spread Assessment — Supplementary Positioning

For known or suspected parotid malignancy, additional targeted sequences through the facial canal are required:

• Coronal T1 fat-suppressed post-contrast through the temporal bone: planned perpendicular to the mastoid portion of the facial nerve (the descending segment of CN VII). This shows the facial canal from the stylomastoid foramen to the geniculate ganglion.

• Axial T1 fat-suppressed post-contrast at the level of the internal auditory canal and CPA cistern: for proximal intracranial extension assessment.

Serial Follow-Up Reproducibility

Document: head coil type; FOV; slice thickness; gap; axial prescription reference (perpendicular to cervical spine at C2–C3 level, superior slice at EAC); coronal prescription reference (parallel to mandibular ramus, centre at parotid isocentre). Reproduce these parameters precisely at follow-up.

Section 4.6 Dedicated Bibliography

Freling NJM, et al. Malignant parotid tumours: clinical use of MR imaging and histologic correlation. Radiology. 1992;185(3):691–696. PMID: 1438747. DOI: 10.1148/radiology.185.3.1438747. (Technical / Foundational) — Foundational MRI-pathological correlation for parotid tumours; establishes slice positioning and T2/T1 characterisation methodology.

Yousem DM, et al. Parotid gland masses: MR imaging characteristics of various tumors. AJNR Am J Neuroradiol. 1989;10(6):1235–1241. PMID: 2511870. (Technical / Foundational) — Early systematic characterisation of parotid tumour MRI appearances; standard reference for positioning and signal characterisation.

5. Optimisation Strategy

5.1 Artifact Reduction by Source

Dental metalwork susceptibility is the dominant artefact challenge in parotid MRI and is not correctable. At 3T, posterior dental amalgam (molars, premolars) produces artefacts that extend 2–5 cm into the deep lobe and parapharyngeal space. Mitigation: 1.5T reduces the artefact radius by approximately 50%; Dixon fat suppression is more robust near metallic artefacts than SPAIR; document the affected region in the report. In patients with bilateral extensive dental metalwork, CT (with soft tissue windows) may be a more reliable modality for assessing the deep lobe medial margin.

Swallowing artefacts: the parotid gland is adjacent to the pharynx, and swallowing artefacts from the adjacent soft palate and pharyngeal walls appear in the A-P phase direction on axial sequences. Mitigation: A-P phase encoding (moves ghosts anteriorly, away from parotid); instruct patient not to swallow during each sequence.

Chemical shift artefact at fat-gland interface: the parotid gland has a characteristic fat-lobular architecture. The fat lobules produce T2 signal. At narrow bandwidth, chemical shift displacement at fat-gland interfaces can produce bright/dark rim artefacts that simulate capsule or mass margin. Wider bandwidth (300–500 Hz/px at 3T) eliminates this.

Flow artefact from retromandibular vein and external carotid artery: the retromandibular vein can produce flow artefact (bright or dark) in the A-P direction on some sequences. This is usually minor and does not interfere with diagnostic assessment; flow compensation (GMN) on T2 sequences reduces this.

Head motion: any head rotation between sequences produces misregistration between the axial T2, T1, and post-contrast T1 — making multiparametric lesion analysis unreliable. Firm head positioning in the coil, and patient instruction to remain still between sequences, reduces this. For restless patients, shorten individual acquisition times.

5.2 Protocol Efficiency and Throughput

A complete parotid MRI — axial T2 non-FS + axial T2-FS + coronal T2-FS + axial T1 + DWI + post-contrast T1-FS axial and coronal — requires approximately 30–40 minutes at 3T.

For a focused single parotid mass characterisation without full bilateral assessment or perineural spread screening: axial T2 non-FS + axial T2-FS + T1 + DWI + post-contrast T1-FS (axial only) ≈ 20–25 minutes.

DCE adds 8–12 minutes and is conditional, reserved for equivocal lesion kinetics or research.

5.3 Field Strength Considerations

3T is preferred for parotid MRI for: (a) higher spatial resolution enabling 0.4–0.5 mm in-plane at 3 mm slice thickness within reasonable acquisition time; (b) improved DWI SNR for ADC characterisation of the 1–2 cm lesions; (c) superior fat-suppressed T2 contrast when SPAIR is effective.

1.5T is preferred or equivalent for: (a) patients with extensive dental metalwork affecting the deep parotid (significantly less susceptibility artefact at 1.5T); (b) patients where the clinical question is adequately answered at lower resolution.

6. Contrast Use Principles Specific to Parotid MRI

6.1 Non-Contrast Standard Protocol — Sufficient For

Non-contrast parotid MRI (T2 non-FS + T2-FS + T1 + DWI) is adequate for: - Classic pleomorphic adenoma characterisation (T2-very bright, T1-dark, well-defined, no restricted diffusion) when clinical and ultrasound context is consistent — surgery can often be planned without contrast in these cases - Classic Warthin tumour characterisation (bilateral, T2-intermediate, restricted diffusion, known clinical context) - Simple intraparotid cyst (T2-very bright, T1-dark, no restriction, no enhancement expected) - Basic gland enlargement characterisation without mass lesion

6.2 Gadolinium Indicated — Region-Specific Contexts

Post-contrast is required or strongly recommended for: - Any suspected malignant parotid tumour: enhancement pattern (heterogeneous enhancement, early washout), perineural spread screening along CN VII, periparotid lymph node enhancement pattern - Parotid abscess: rim enhancement vs diffuse phlegmon; deep space extension - Equivocal mass lesion: when T2 signal and ADC are not diagnostic (malignant vs Warthin tumour overlap), dynamic enhancement kinetics provide additional differentiation - Post-surgical assessment for recurrence: enhancement distinguishes post-surgical scar (non-enhancing mature scar) from viable recurrent tumour - Sjögren syndrome with suspected lymphomatous transformation: focal enhancement in a Sjögren gland suggests lymphoma - Any CN VII palsy associated with parotid disease: perineural spread detection requires post-contrast

6.3 Post-Contrast Acquisition Timing

Standard equilibrium phase (3–5 minutes post-injection) for axial and coronal T1-FS post-contrast provides adequate characterisation of most parotid pathology.

DCE kinetic analysis (when performed): typically acquired as 8–12 consecutive phases at 40–60 seconds per phase, starting from injection. The time-intensity curve (TIC) type — as in breast MRI — categorises enhancement as: type I (persistent = benign, typical of pleomorphic adenoma); type II (plateau); type III (washout = malignant or Warthin tumour). DCE for parotid requires the same temporal resolution as breast DCE and is implemented as a 3D T1 fat-suppressed sequence with the DCE parameters described in the Breast MRI master page.

Delayed post-contrast (8–15 minutes): for perineural spread and facial canal assessment, delayed post-contrast T1 (at 8–15 minutes) provides persistent or increasing enhancement along the nerve that may be more conspicuous than early-phase enhancement.

7. Reporting Essentials

7.1 Interpretation Framework

Parotid MRI reporting follows a systematic approach:

Gland survey: bilateral size comparison; parenchymal T2 signal homogeneity (homogeneous = normal; heterogeneous = tumour, inflammation); fat lobule pattern (preserved = benign; replaced = malignant or diffuse disease).

Mass characterisation (if present): 1. Location: superficial vs deep lobe (relative to retromandibular vein); parotid tail; accessory parotid 2. Size (three dimensions) 3. T2 signal: very bright (cystic/myxoid/pleomorphic adenoma) / intermediate (Warthin, malignant) / dark (fibrosis, calcification, haemosiderin) 4. T1 signal: dark / intermediate / bright (fat, haemorrhage) 5. Margin: well-defined/lobulated (benign) / ill-defined/irregular (malignant) 6. Internal architecture: homogeneous / heterogeneous / necrotic / cystic 7. ADC value (if DWI acquired) 8. Enhancement pattern (if contrast administered)

CN VII assessment: asymmetric enhancement, thickening, or T2 signal change along the expected facial nerve course; stylomastoid foramen signal; facial canal (mastoid and tympanic segments); geniculate ganglion.

Lymph nodes: intraparotid (level Ib), periparotid, level II — size, morphology, necrosis, ADC.

Deep lobe and parapharyngeal space: parapharyngeal fat intact/obliterated; retromandibular vein position; deep lobe extension.

Broad axes: benign vs malignant (margin, T2 signal, ADC, enhancement kinetics); focal vs diffuse; unilateral vs bilateral.

7.2 Mandatory Reporting Checklist

Technical quality: - [ ] Field strength; coil type - [ ] Fat suppression quality: SPAIR/STIR uniform / failure (region affected; dental metalwork artefact documented) - [ ] Head position: symmetric / asymmetric (rotation noted)

Right parotid gland: - [ ] Size (cm) - [ ] Parenchymal signal: normal / abnormal - [ ] Mass: present / absent - If present: location (superficial/deep/tail); size; T2 signal; T1 signal; margin; ADC; enhancement - Retromandibular vein: preserved / displaced / compressed / obscured - Deep lobe: not involved / involved; parapharyngeal fat: intact / obliterated - [ ] Intraparotid lymph nodes: normal / abnormal

Left parotid gland: same checklist

CN VII (if relevant): - [ ] Stylomastoid foramen: normal / T2 signal change / enhancement - [ ] Facial canal: normal / asymmetric enhancement / widening

Periparotid and level I–II nodes: normal / abnormal (size, necrosis, ECS)

Adjacent structures: mandible / temporal bone: normal / involved; masseter: normal / invaded

7.3 Structured Reporting

Reports must include: Indication; Technique (field strength, coil, sequences, contrast: agent, dose, timing); Comparison; Findings (bilateral systematic review); Impression (with suggested differential diagnosis if mass present: BI-RADS equivalent classification of malignant risk); CN VII statement (for parotid mass: “No evidence of perineural spread along CN VII to the level of the geniculate ganglion / stylomastoid foramen”); Recommendations (ultrasound-guided FNAC; further staging if malignancy suspected; follow-up interval if benign); Limitations (dental metalwork, fat suppression failure region).

7.4 Incidental Findings — Clinical Decision Framework

Usually benign: small intraparotid lymph node with fatty hilum (< 10 mm short-axis); simple intraparotid cyst (T2-very bright, no wall, no enhancement); bilateral lymphoepithelial cysts (multiple bilateral T2-bright cysts in HIV context — incidental finding); fat lobules within the gland (normal adult parenchyma).

Follow-up required: incidental parotid mass without clear benign features — recommend ultrasound-guided FNAC; stable small mass with classic pleomorphic adenoma features on a prior study — repeat MRI in 12 months if surgery deferred.

Urgent communication required: unexpected facial nerve canal enhancement suggesting perineural spread from an unsuspected malignancy; unexpected bilateral enlargement with T2-dark masses suggesting lymphoma; intracranial extension of parotid tumour; unexpected carotid space involvement.

8. MRI Technologist Pearls

8.1 Sequence Order Logic

1. Three-plane localiser ← verify bilateral symmetric head positioning
2. Axial T2 non-FS ← anatomical map; retromandibular vein identification; early in case of patient fatigue
3. Axial T2-FS (SPAIR) ← pathological signal
4. Coronal T2-FS ← bilateral overview and nodal chains
5. Axial T1 ← T1 characterisation; pre-contrast reference
6. DWI ← pre-contrast; ADC characterisation
7. Contrast injection (if indicated)
8. Axial T1-FS post-contrast ← enhancement characterisation
9. Coronal T1-FS post-contrast ← enhancement bilateral; perineural screening
10. DCE (if indicated — after standard post-contrast if kinetics required)

8.2 Positioning Tricks

For patients with large ears or prominent tragal cartilage: the head coil cup may compress the ear against the temporal bone, producing patient discomfort and head movement. Thin foam ear protection (not hearing protection, just soft foam) between the ear and the coil side wall allows comfortable positioning.

For patients with a unilateral parotid mass: if the mass is on the right side, centring the head very slightly rightward (1–2 cm) within the coil improves coil coupling on the affected side. This is only appropriate for very large unilateral tumours where maximising SNR on the target side is justified.

For patients with dentures: dentures should be kept in situ if they are close-fitting and not metallic (plastic dentures produce minimal artefact). Removing loose-fitting dentures changes the oral cavity volume and may cause jaw movement during the examination.

8.3 Fast Salvage Protocol

Approximately 11 minutes — covers T2 characterisation + enhancement. Adequate for most parotid mass referrals. ADC cannot be assessed without DWI; CN VII perineural spread cannot be assessed without coronal sequences.

8.4 Common Avoidable Errors

9. Quality Control Checklist

• Head position symmetric: mandibular condyles at same height on axial localiser
• EAC included in superior coverage (bilateral)
• Mandibular angle included in inferior coverage (bilateral)
• Axial T2 non-FS: retromandibular vein visible bilaterally
• Axial T2-FS: fat suppression quality assessed — uniform / failure (dental metalwork: document level)
• DWI: ADC map generated; slice thickness ≤ 4 mm
• Post-contrast T1-FS: both axial AND coronal acquired if contrast given
• Coronal T1-FS: stylomastoid foramen level included (for malignant mass cases)
• All sequences pre-contrast STIR acquired before gadolinium injection
• Correct laterality labelling: right and left parotid confirmed on axial images
• Comparison with prior imaging or ultrasound performed before reporting

11. Evidence Gaps and Ongoing Debate

ADC threshold for malignant vs benign parotid masses: published ADC cutoffs for distinguishing malignant from benign parotid tumours range from 0.9 to 1.4 × 10⁻³ mm²/s across studies. The overlap between Warthin tumour (low ADC, benign) and malignant parotid tumours (also low ADC) significantly limits the specificity of DWI for malignancy detection. No universally validated single threshold exists.

DCE kinetic analysis clinical utility: the TIC classification (types I, II, III) for parotid mass differentiation has been described in multiple series with promising sensitivity and specificity data. However, methodological differences — DCE protocol (temporal resolution, FA, injection timing), ROI placement (inside lesion vs hotspot), and thresholds used — vary across studies, preventing definitive guideline recommendations.

3T vs 1.5T for parotid: while 3T provides nominally higher spatial resolution, the increase in dental metalwork artefact at 3T remains a significant barrier in clinical practice where patients frequently have posterior dental restorations. No formal prospective comparison of diagnostic accuracy for parotid tumour characterisation at 3T vs optimised 1.5T exists.

MR sialography standardisation: various protocols for stimulated ductal secretion (lemon juice, citric acid, pilocarpine) combined with heavily T2-weighted sequences have been described for parotid duct assessment in Sjögren and sialolithiasis. No standardised MR sialography protocol has been adopted across platforms.

AI-assisted parotid lesion classification: deep learning algorithms for parotid mass classification (benign vs malignant; histological subtype prediction) have been described in single-institution studies with accuracy 80–90%. No FDA/EMA-cleared clinical tool exists.