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

1. Executive Summary

Brachial plexus MRI is one of the most technically demanding routine MRI examinations in clinical practice. The brachial plexus — formed by the ventral rami of C5 through T1, coursing through the intervertebral foramina, the posterior triangle of the neck, the scalene triangle, and the axilla — consists of neural elements that measure 2–5 mm in diameter, run obliquely through multiple tissue planes, cross the cervicothoracic junction (where B0 homogeneity is challenging), and are surrounded by fat, muscle, vessels, and bone in close proximity. The diagnostic challenge is to visualise these small structures continuously from the spinal cord to the axillary nerves, to detect signal abnormality within them, and to identify adjacent pathology compressing or invading them — all at adequate spatial resolution, with reliable fat suppression, and without prohibitive acquisition time.

MRI is the modality of choice for all non-acute brachial plexus assessment. It supersedes CT in soft tissue and neural characterisation; it supersedes ultrasound in the ability to image the pre- and postfixed plexus from the foraminal level through the retroclavicular region (which is sonographically inaccessible); and it is the only modality that can visualise the nerve roots at the spinal cord level, detect pseudomeningoceles (avulsion injuries), and characterise perineural infiltration by tumour. The ACR Appropriateness Criteria [1] designate MRI as the primary imaging modality for brachial plexus neuropathy, trauma, and mass lesion assessment.

The generic brachial plexus protocol provides a comprehensive bilateral assessment of the full plexus from the spinal cord to the axillary level, using coronal and axial planes with dedicated fat suppression and — in modern practice — neurographic sequences optimised for nerve contrast. This generic protocol is the starting point for all brachial plexus indications; specific clinical questions (obstetric palsy, Pancoast tumour infiltration, post-radiation plexopathy) require targeted protocol modifications described at the child page level.

1.1 Core Strengths

Complete nerve tract visualisation: MRI is the only modality that visualises the brachial plexus continuously from the spinal cord roots (C5–T1 root sleeves, cord-root interface) through the foramina, the pre- and postfixed plexus, and to the axillary level — a span of approximately 20 cm. Ultrasound cannot image the retroclavicular and foraminal components; CT cannot characterise neural signal.

Pseudomeningocele detection: avulsion of the nerve root from the spinal cord produces tearing of the root sleeve and meningeal layers, allowing CSF to escape into the epidural and paraspinal spaces — forming a pseudomeningocele. On MRI (heavily T2-weighted sequences), pseudomeningoceles appear as CSF-bright sacs adjacent to the expected root location. They are pathognomonic for preganglionic (avulsion) injury and indicate non-reconstructable damage. CT myelography historically provided this information; MRI with heavily T2-weighted sequences (CISS/DRIVE or 3D constructive interference in steady state) has replaced CT myelography in most centres.

Neural signal characterisation: abnormal T2 signal within nerve trunks — T2 hyperintensity from oedema in neuritis, tumour infiltration, post-radiation change — is visible on fat-suppressed T2 sequences. The normal nerve is T2-isointense to muscle; any nerve that is brighter than muscle on STIR or fat-suppressed T2 is abnormal. This T2 neural signal change is invisible on CT and ultrasound.

Mass lesion characterisation: nerve sheath tumours (schwannoma, neurofibroma, plexiform neurofibroma in NF1), metastatic infiltration, and reactive nodal masses adjacent to the plexus are fully characterised on MRI. Schwannomas are T2-bright with a target sign (dark central area, bright peripheral zone) on high-resolution coronal T2. Neurofibromas show similar T2 features but are fusiform and non-eccentric. Perineural tumour spread along the plexus produces T2 signal increase and enhancement tracking along the nerve.

Pre- vs postganglionic distinction: this is the most clinically critical application. Preganglionic injury (root avulsion, at or proximal to the dorsal root ganglion) cannot be surgically repaired; postganglionic injury (distal to the DRG) can potentially be reconstructed. MRI reliably identifies pseudomeningoceles (preganglionic marker) and assesses root continuity at the foraminal level.

Bilateral simultaneous assessment: both plexuses are imaged simultaneously in the coronal plane, enabling direct comparison of the normal side with the abnormal side — the most reliable method for identifying unilateral abnormality given the complex anatomy and individual variation.

1.2 Intrinsic Limitations of the Generic Protocol

Small structure size at the spatial resolution limit: the brachial plexus nerve elements at the cervicothoracic junction measure 2–5 mm. At clinical 1.5T–3T with a spine or neck/body coil combination, achieving the required 0.8–1.5 mm in-plane resolution across the full bilateral plexus FOV (typically 30–40 cm width) within a reasonable acquisition time requires aggressive parallel imaging and partial Fourier acquisition, which reduces SNR. Small nerve signal changes may be at the detection threshold.

B0 inhomogeneity at the cervicothoracic junction: the brachial plexus crosses the cervicothoracic junction where the lungs, the subclavian vessels, and the first rib produce major B0 field disturbances. This is one of the most challenging B0 environments in the entire body for fat suppression — SPAIR/CHESS fail predictably, particularly at the level of the first rib and the scalene triangle. STIR is the mandatory fat suppression strategy for this reason.

Retroclavicular and infraclavicular plexus: the plexus posterior to the clavicle and through the axilla is partially obscured by susceptibility from the clavicle, the first rib, and axillary vessels. Coverage of the infraclavicular plexus (lateral cord, medial cord, posterior cord, and terminal branches) requires additional axial sequences centred on the axilla, which may extend scan time substantially.

The generic protocol does not fully cover the spinal cord: the cervical spinal cord pathology that may coexist with brachial plexus lesions (syringomyelia, spinal cord tumour, central cord injury) requires a dedicated cervical spine protocol. The generic brachial plexus protocol provides limited cord coverage and is not designed as a spinal cord assessment tool.

When dedicated child protocols are required: obstetric brachial plexus palsy (requires specific foraminal sequences and DRG characterisation); Pancoast tumour plexus infiltration (requires chest MRI integration); post-radiation plexopathy vs recurrence (requires contrast and comparison with post-treatment baseline); nerve sheath tumour characterisation (NF1: requires full body neurography context); traumatic plexopathy with surgical planning (requires functional MRI neurography); thoracic outlet syndrome (requires dynamic sequences); post-operative plexopathy assessment.

2. Main Clinical Indications

2.1 Standard Indications

Traumatic brachial plexus injury is the most clinically urgent indication. High-velocity trauma (motorcycle accidents, falls from height, difficult deliveries) produces a spectrum of injury from neuropraxia to root avulsion. The most critical information MRI provides: (a) identification of pseudomeningoceles indicating root avulsion (preganglionic, non-reconstructable); (b) assessment of root continuity at the foraminal level; (c) paraspinal muscle denervation oedema as an indirect preganglionic injury marker (STIR-bright paraspinal muscles ipsilateral to avulsion). The generic protocol covers all of these. For surgical planning requiring selective nerve root identification, dedicated high-resolution sequences (CISS/DRIVE for root sleeves and pseudomeningoceles) supplement the generic protocol.

Brachial plexus neuropathy (Parsonage-Turner syndrome / neuralgic amyotrophy) presents with acute severe shoulder and upper arm pain followed by motor weakness. MRI demonstrates T2 signal increase in the affected nerve trunks/divisions and, on STIR, denervation oedema in the muscles supplied by the affected nerves. The generic protocol is adequate for diagnosis. Serial MRI at 3–6 months demonstrates evolving denervation atrophy.

Cervical rib and thoracic outlet syndrome: a cervical rib or C7 transverse process elongation compresses the lower trunk of the brachial plexus (C8–T1). MRI demonstrates: the rib or elongated transverse process (seen on coronal T1 as a bony structure inferior to the expected C7 level); displacement or compression of the lower trunk; T2 signal change in the lower trunk. The generic coronal protocol is the starting point; dynamic sequences (arm positioning) may be added for suspected vascular thoracic outlet syndrome.

Mass lesion adjacent to or involving the brachial plexus: nerve sheath tumours (schwannoma, neurofibroma), lipomatous lesions (lipoma, lipofibromatous hamartoma), vascular malformations, and metastatic infiltration. The generic protocol with coronal STIR and T1 + post-contrast T1 is adequate for initial characterisation. Dedicated neurography sequences (3D MRN, DWIBS) improve nerve delineation for surgical planning.

Pancoast tumour (superior sulcus tumour): apical lung tumours invading the first rib, subclavian artery, and lower trunk of the brachial plexus. MRI is the definitive modality for assessing the extent of T1 infiltration and plexus involvement — critical for surgical resectability assessment. The generic brachial plexus protocol combined with lung apex imaging (requires axial coverage into the superior lung) provides this information. A dedicated Pancoast protocol integrating chest MRI sequences is the optimal approach.

Post-radiation plexopathy: after radiotherapy to the breast, axilla, or neck/mediastinum, radiation plexopathy presents months to years later as progressive sensorimotor deficit. MRI distinguishes radiation fibrosis (T2-dark, non-enhancing, diffuse) from recurrent tumour (T2-bright, enhancing, focal) — a critically important distinction that changes management. The generic protocol with contrast is appropriate; baseline post-radiation MRI is the most important comparator.

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

Analgesic management: brachial plexus injuries and neuropathies are frequently associated with severe neuropathic pain. Positioning the affected arm for 30–45 minutes may be intolerable without adequate analgesia. Co-ordinate pre-examination analgesia (oral analgesic 30–60 minutes before the examination) with the referring team. For post-traumatic patients with associated shoulder or rib injuries, this is particularly important.

Arm positioning options: the standard bilateral brachial plexus MRI is performed with the arms alongside the body. This is the most comfortable position and reduces motion from shoulder fatigue. Alternative: arms slightly abducted (30°) with slight external rotation reduces susceptibility from the first rib and clavicle near the plexus. For suspected thoracic outlet syndrome, specific dynamic arm positions (arms elevated, Adson manoeuvre) may be required as conditional sequences.

Metallic implants near the plexus: prior shoulder surgery (clavicle plating, acromioclavicular fixation, pacemaker leads near the subclavian) and cervical spine hardware (fusion cages, anterior cervical plating) produce susceptibility artefacts that may obscure portions of the plexus. At 3T, the artefact radius is substantially larger than at 1.5T; for patients with extensive metallic implants near the plexus, 1.5T is preferred.

Pacemakers and cardiac implantable electronic devices: CIEDs require specific compatibility assessment before brachial plexus MRI — this is a proximity examination and the device leads may course near the imaging volume. Refer to the general preparation page protocol for CIED management.

Brachial plexus anatomy variant history: unilateral prominent C4 contribution (prefixed plexus) or T2 contribution (postfixed plexus) changes the expected anatomy. If electrodiagnostic (EMG/NCS) data is available, review it before the examination to understand which roots and trunks are abnormal — this guides the radiologist's interpretation and may guide targeted sequence planning.

3.2 Patient Positioning on the MRI System

Position: supine, head-first, arms alongside the body. The neutral arm position reduces motion and is comfortable for most patients. A small pillow under the head maintains neutral neck alignment.

Coil selection: the brachial plexus requires a coil combination that provides adequate SNR across the full span from C4 to the axilla bilaterally — a craniocaudal span of approximately 20–25 cm and a lateral span of approximately 35–45 cm.

Options:

• Head-and-neck + body matrix coil combination: the posterior spine coil elements at the cervicothoracic level combined with the anterior neck coil and body matrix provide the most complete coverage. This is the standard configuration for most departments.
• Spine coil + shoulder phased-array coil (ipsilateral): for unilateral focused plexus assessment (post-traumatic or neoplastic), the shoulder coil placed over the ipsilateral supraclavicular fossa and axilla provides superior SNR for the target side.
• Large flexible phased-array: some departments use a large flexible coil wrapped around the neck and both shoulders for bilateral assessment; SNR is lower than dedicated spine+neck coils but coverage is complete.

Centring: isocentre at the level of the cervicothoracic junction — approximately at the level of the sternoclavicular joints (T1–T2 level). Verify on the three-plane localiser that: the C4–C5 level (upper plexus) is within the superior coverage; the axillary level (terminal branches) is within the inferior coverage; both clavicles are visible bilaterally.

Head position: neutral, with a head support maintaining symmetric alignment. Rotation of the head produces asymmetric coil coupling and asymmetric fat suppression between the two sides — a bilateral comparison examination requires symmetric positioning.

Arm position: arms extended alongside the body, palms facing down. Verify that neither arm is externally rotated (which moves the bicipital groove and produces asymmetric shoulder position) or internally rotated (which brings the humeral head closer to the plexus on one side, producing susceptibility asymmetry).

Technologist verification before starting: (a) both clavicles visible on coronal localiser; (b) C4 foramen visible on sagittal localiser; (c) axillary content visible at inferior margin; (d) isocentre at sternoclavicular joint level.

4. Standard Protocol Design

4.1 Mandatory Core Sequences

4.2 Conditional Sequences

4.3 Rationale Summary Per Sequence

Coronal T1 non-fat-suppressed is the anatomical reference sequence for the brachial plexus and is the sequence on which the plexus anatomy is most clearly displayed in its entirety. On coronal T1: the nerve trunks of the brachial plexus appear as intermediate-signal tubular structures within the surrounding bright fat of the posterior triangle and scalene triangle; the background fat provides natural contrast that outlines each trunk. The fat signal is preserved (not suppressed) specifically to create this nerve-versus-fat contrast. The T1 coronal provides: (a) bilateral plexus anatomy in a single acquisition; (b) identification of space-occupying lesions — T1-dark masses within the fat-containing plexus spaces; (c) the anatomical landmarks for planning all other sequences; (d) bone detail (cervical spine, first rib, clavicle, humerus head).

Coronal STIR is the primary pathology detection sequence. On STIR: the fat background is nulled; muscles are intermediate signal; the nerve trunks appear as intermediate-to-slightly-bright structures (nerve T2 slightly longer than muscle); pathological signal change (T2-prolonged nerve from oedema, infiltration, demyelination) appears as discrete hyperintensity distinguishable from the dark fat background. STIR coronal is the most sensitive sequence for: (a) T2 signal change within nerve trunks; (b) paraspinal muscle denervation oedema (STIR-bright muscle ipsilateral to nerve root avulsion — particularly the paraspinal muscles at the C5–T1 levels, which are denervated when the roots are avulsed); (c) perineural soft tissue oedema from adjacent tumour or haematoma.

The STIR contraindication post-gadolinium applies here as universally throughout MRIninja: STIR must be acquired before any gadolinium injection.

Axial T2-FS (STIR or SPAIR) provides cross-sectional anatomy of the plexus elements at each level. The axial sequences are essential for: (a) understanding the spatial relationship of the plexus to adjacent structures at each level (scalene muscles, subclavian artery and vein, cervical spine); (b) identifying focal nerve enlargement at a specific level; (c) characterising adjacent masses (retroclavicular mass, lymph node vs nerve sheath tumour vs vascular lesion); (d) assessing the infraclavicular plexus (a dedicated axillary-level axial sequence may be needed). The axial T2-FS is also the most useful sequence for detecting early post-radiation nerve T2 signal change in the cross-sectional view.

Axial T1 provides the T1 characterisation of cross-sectional findings and allows assessment of the bone marrow signal in the cervical vertebrae (relevant for metastatic disease from Pancoast or breast tumour).

3D heavily T2-weighted (CISS/DRIVE/FIESTA-C): this is the sequence that has most changed brachial plexus MRI over the past decade. Using a 3D balanced steady-state free precession (bSSFP) readout with very long effective TE (CISS — Constructive Interference in Steady State; GE: FIESTA-C; Philips: DRIVE; Canon: FASE), the sequence produces extremely high fluid-to-tissue contrast — CSF appears very bright; nerve roots appear as intermediate-dark structures within the bright CSF in the spinal canal; pseudomeningoceles appear as CSF-bright sacs outside the expected root location. At sub-millimetre isotropic resolution (0.7–1.0 mm), the individual rootlets C5–T1 are visible within the subarachnoid space, enabling: (a) identification of absent root — the empty root sleeve sign of complete avulsion; (b) pseudomeningocele as a pathognomonic avulsion marker; (c) assessment of root sleeve symmetry bilaterally; (d) detection of intradural pathology (spinal cord involvement). This sequence has replaced CT myelography for pre-ganglionic injury assessment in most expert centres [3, 4].

4.4 Sequence Matching and Cross-Sequence Consistency

The coronal T1 and coronal STIR must use identical slice geometry (same level positions, same FOV, same slice thickness and gap) to enable direct visual comparison of the same anatomical level in T1 and STIR. This is particularly important for confirming that a T1-bright structure (fat or haemorrhage) corresponds to what appears STIR-suppressed or STIR-bright on the companion sequence.

The 3D CISS/DRIVE acquisition provides isotropic resolution and can be reformatted in any plane. Reformatted oblique coronal views along the nerve root trajectory provide the most complete root sleeve analysis.

For serial examinations (post-treatment, treatment response), geometric reproducibility is essential. Document the coronal angulation relative to the C7–T1 disc level on the sagittal localiser; reproduce at each follow-up.

4.5 Fat Suppression — Brachial Plexus-Specific

STIR is the mandatory fat suppression technique for brachial plexus MRI — not SPAIR, not CHESS. The rationale is identical to the WB-MRI for myeloma and the neck soft tissues protocol: the cervicothoracic junction produces massive B0 inhomogeneity from the lung apex-mediastinum-first rib interfaces. At this level, spectral fat saturation (SPAIR/CHESS) fails reliably, producing residual bright fat signal in the lower plexus region and scalene triangle — precisely where pathological nerve T2 signal is most clinically important. STIR fat suppression is B0-field-independent and provides uniform suppression from the skull base to the axilla regardless of local B0 variation.

The only exception: post-contrast T1 fat-suppressed sequences use Dixon or SPAIR (not STIR). STIR post-gadolinium is an absolute contraindication as documented throughout the MRIninja protocol series.

Dixon for post-contrast T1: at 3T, Dixon fat-water separation provides B0-independent fat suppression for post-contrast T1 sequences, directly addressing the B0 challenge at the cervicothoracic junction.

5. Optimisation Strategy

5.1 Artifact Reduction by Source

B0 inhomogeneity fat suppression failure at cervicothoracic junction: as discussed, this is the dominant technical limitation of brachial plexus MRI. The lung apex, first rib, subclavian vessels, and mediastinum produce massive local B0 field variations at precisely the level where the lower trunk (C8–T1) is most clinically important. STIR provides the only reliable fat suppression at this level. Even with STIR, partial fat suppression failure may occur at the extreme apex of the lung on the anterior surface of the plexus — document this in the report when present.

Respiratory motion artefacts: the brachial plexus examination does not require breath-holding (the plexus itself is not a respiratory organ). However, respiratory motion of the chest wall and lung apex produces artefacts that propagate in the phase-encoding direction. R-L phase encoding (coronal) displaces these artefacts out of the bilateral plexus path. For coronal sequences, respiratory triggering is not standard but may be applied when respiratory motion is severe.

Vascular pulsation artefacts: the subclavian artery and the common carotid artery produce pulsatile artefacts in the phase-encoding direction. Saturation bands placed superior and inferior to the imaging volume reduce the inflowing blood signal that contributes to pulsation artefacts. Flow-compensated (GMN) sequences reduce vessel-related T2 signal artefacts.

Susceptibility from the first rib and clavicle: the cortical bone of the first rib and clavicle produces local susceptibility that causes signal dropout in the adjacent plexus elements on GRE-based sequences. This affects the 3D CISS/FIESTA-C (bSSFP) sequence, which is susceptibility-sensitive. At the level of the first rib and clavicle, signal dropout on 3D CISS may limit assessment of the middle trunk and lower trunk as they cross this level. Increasing the TE (TE + = longer time for susceptibility dephasing) exacerbates the problem; using the shortest achievable TE in the bSSFP sequence minimises it.

Swallowing artefacts: as in the neck soft tissue protocol, swallowing produces artefacts in the anterior neck. The brachial plexus is more lateral and posterior than the larynx, so swallowing artefacts are less severe; however, they are still visible in the anterior triangle on coronal sequences. Patient instruction (breathe through nose; do not swallow during sequences) reduces but does not eliminate this artefact.

Dental metalwork: as in the neck protocol, dental metalwork produces susceptibility artefacts in the upper plexus (C5–C6) region in the posterior triangle of the neck at the level of the mandible. Document if present.

5.2 Protocol Efficiency and Throughput

A complete brachial plexus MRI — coronal T1 + coronal STIR + axial T2-FS + axial T1 + 3D CISS + post-contrast T1-FS (when indicated) — requires approximately 40–55 minutes at 3T.

For a focused assessment (e.g., suspected upper trunk injury): coronal T1 + coronal STIR + 3D CISS ≈ 20–25 minutes. This abbreviated approach is clinically adequate for initial traumatic or Parsonage-Turner evaluation.

For post-contrast assessment (neoplastic plexopathy): add coronal and axial T1-FS post-contrast: approximately 10 minutes extra.

The 3D CISS/DRIVE sequence adds 6–10 minutes but provides the pseudomeningocele information that cannot be obtained from any other sequence — always include it in traumatic and surgical planning contexts.

5.3 Field Strength Considerations

3T is preferred for brachial plexus MRI for: (a) superior SNR enabling 0.8–1.0 mm in-plane resolution for the small nerve trunks; (b) higher SNR for the 3D CISS sequence enabling sub-millimetre isotropic resolution; (c) better DWI quality for DWIBS neurography.

Key 3T challenge: the bSSFP banding artefacts that affect 3D CISS/FIESTA-C at 3T are substantially more severe than at 1.5T. The bSSFP sequence relies on B0 homogeneity for off-resonance suppression; at 3T, off-resonance banding artefacts may obscure the nerve root assessment in the cervicothoracic region near susceptibility sources (first rib). DRIVE (Philips variant using RARE readout instead of bSSFP) and some SPACE variants avoid bSSFP banding. At 1.5T, bSSFP banding is less severe and the 3D CISS is generally more uniform across the FOV.

1.5T preferred for: (a) patients with extensive metallic hardware near the plexus; (b) when 3T bSSFP banding artefacts are unacceptable at the cervicothoracic junction.

6. Contrast Use Principles Specific to Brachial Plexus MRI

6.1 Non-Contrast Standard Protocol — Sufficient For

Non-contrast brachial plexus MRI (coronal T1 + coronal STIR + axial T2-FS + axial T1 + 3D CISS) is diagnostically adequate for:

• Traumatic brachial plexus injury assessment (pseudomeningocele detection; denervation oedema; root continuity)
• Parsonage-Turner syndrome (T2 signal change in nerve trunks; denervation pattern)
• Cervical rib and thoracic outlet compression (nerve displacement; T2 signal change in lower trunk)
• Nerve sheath tumour initial characterisation (T2 signal pattern; size; anatomical level)
• Anatomical variant assessment
• Most post-traumatic cases

6.2 Gadolinium Indicated — Region-Specific Contexts

Post-contrast T1-FS sequences are required or strongly useful for:

• Suspected neoplastic infiltration (Pancoast tumour, breast cancer recurrence, lymphoma): enhancement of the infiltrating tumour within the plexus; perineural spread along the nerve trunk
• Nerve sheath tumour characterisation: homogeneous enhancement in schwannoma; heterogeneous in large or malignant nerve sheath tumour (MPNST); enhancement pattern distinguishes benign from malignant
• Post-radiation plexopathy vs recurrent tumour: radionecrosis/fibrosis — T2-dark, non-enhancing; tumour recurrence — T2-bright, enhancing
• Infectious or inflammatory plexitis: periNeural enhancement in reactive neuritis; abscess enhancement pattern
• Parsonage-Turner syndrome with diagnostic uncertainty: nerve enhancement may be present in acute neuritis; enhancement distinguishes acute active inflammation from chronic atrophy

6.3 Post-Contrast Acquisition Timing

Standard equilibrium phase (3–5 min post-injection) for coronal and axial T1-FS post-contrast. No specific early or delayed phase protocol is required for the generic brachial plexus indication. Document injection time.

7. Reporting Essentials

7.1 Interpretation Framework

Brachial plexus MRI reporting requires systematic analysis of the plexus at each anatomical level from roots to terminal branches:

Level-by-level analysis:

• Roots (C5–T1 at the foraminal level): are all roots present bilaterally? Symmetric foraminal fat signal? Pseudomeningocele present?
• Trunks (upper = C5+C6; middle = C7; lower = C8+T1): size, T2 signal (compare to ipsilateral normal trunk or contralateral), enhancement if contrast given
• Divisions (in the retroclavicular space): less individually assessable but the transition from trunks to divisions is a recognisable level
• Cords (lateral, medial, posterior — at the axillary level): cord anatomy and signal; requires targeted axillary-level imaging

Broad diagnostic axes:

• Pre- vs postganglionic (pseudomeningocele = preganglionic; absent pseudomeningocele with foraminal T2 change = postganglionic)
• Traumatic vs neoplastic vs inflammatory vs radiation (clinical context + T2 signal pattern + enhancement)
• Focal vs diffuse signal change (focal = traumatic, tumour; diffuse = radiation, Parsonage-Turner)
• Unilateral vs bilateral (unilateral = traumatic, neoplastic; bilateral = inflammatory, radiation, thoracic outlet)
• Compressive vs intrinsic neural signal change (compressive = extrinsic mass, rib, vessel; intrinsic = avulsion, neuritis, tumour)

Muscle denervation patterns: the muscles of the upper extremity have predictable innervation patterns. STIR-bright signal in specific muscle groups — appearing within 1–2 weeks of acute denervation, persisting for months — indicates damage at the corresponding nerve level:

• Paraspinal muscles (deep cervical): root avulsion (preganglionic)
• Rhomboids (dorsal scapular nerve): C5 root avulsion (proximal to rhomboid branch)
• Serratus anterior (long thoracic nerve, C5–C7): upper/middle trunk
• Supraspinatus/infraspinatus (suprascapular nerve, C5–C6): upper trunk
• Deltoid (axillary nerve, C5–C6): upper trunk/posterior cord

7.2 Mandatory Reporting Checklist

Technical quality:

• Field strength; coil type
• STIR fat suppression: uniform / partial failure (level affected)
• Motion artefacts: acceptable / moderate (affected sequence)
• Bony implant artefacts: absent / present (affected region)

Roots (bilateral, C5–T1):

• Each root sleeve: present / absent / enlarged / signal change
• Pseudomeningocele: absent / present (level, size, ipsilateral)
• Foraminal fat: normal / replaced
• Foramen ovale / neural canal: patent / narrowed

Trunks (bilateral):

• Upper trunk (C5–C6): normal / T2 change / enlarged
• Middle trunk (C7): normal / T2 change
• Lower trunk (C8–T1): normal / T2 change
• Size: symmetric / asymmetric (mm if measured)

Retroclavicular and axillary components:

• Accessible / obscured by susceptibility
• Cord level: assessable / not assessable on this protocol

Muscles — denervation assessment:

• Paraspinal muscles: normal / STIR signal / atrophy
• Periscapular muscles (rhomboids, serratus): normal / denervation
• Rotator cuff muscles: normal / denervation
• Biceps/triceps: normal / denervation (specify side)

Adjacent structures:

• Cervical vertebrae / first rib: normal / abnormal (fracture, mass, bone marrow change)
• Subclavian artery/vein: patent / compressed / thrombosed
• Lymph nodes: normal / pathological

Post-contrast (if acquired):

• Nerve trunk enhancement: absent / present (level, pattern)
• Adjacent mass enhancement

7.3 Structured Reporting

Reports must include: Indication (trauma / neuropathy / tumour / post-radiation); Technique (field strength, coil, sequences, 3D CISS used, contrast if given); Comparison (prior brachial plexus MRI, EMG/NCS data); Findings (systematic by anatomical level); Pre- vs postganglionic assessment (for traumatic cases); Impression (answer to clinical question; level and extent of injury/pathology; surgical implications); Limitations (fat suppression failure region, susceptibility from hardware, coverage gaps).

7.4 Incidental Findings — Clinical Decision Framework

Usually benign: small cervical rib (noted as anatomical variant if asymptomatic); mild C-spine degenerative disease without foraminal compromise; small incidental nerve sheath tumour (schwannoma < 1 cm, T2-target sign, no enhancement asymmetry, stable on prior).

Follow-up required: isolated mild T2 signal change in a single trunk without clinical symptoms (reassess with clinical correlation); mildly enlarged lymph node adjacent to plexus without features of metastasis; isolated unilateral muscle denervation signal without electrophysiological correlation.

Urgent communication: pseudomeningocele at multiple levels (total avulsion — surgical urgency); unsuspected Pancoast tumour involving plexus; cord signal change from adjacent tumour or disc; subclavian artery occlusion; new pathological vertebral fracture; unexpected bilateral plexopathy suggesting paraneoplastic or systemic process.

8. MRI Technologist Pearls

8.1 Sequence Order Logic

1. Three-plane localiser ← verify positioning; confirm bilateral plexus coverage
2. Coronal T1 ← anatomical map; most diagnostic for plexus anatomy; early if patient fatigues
3. Coronal STIR ← pathology detection; before contrast
4. Axial T2-FS (STIR) ← cross-sectional anatomy
5. Axial T1 ← T1 characterisation
6. 3D CISS/DRIVE ← root sleeve and pseudomeningocele assessment
7. Contrast injection (if indicated)
8. Coronal T1-FS post-contrast
9. Axial T1-FS post-contrast

The coronal T1 is acquired first — it is the most anatomically informative sequence and is relatively motion-insensitive. If the patient cannot complete the full protocol, having the coronal T1 and coronal STIR provides the essential diagnostic information.

8.2 Positioning Tricks

For patients with suspected lower trunk compression by cervical rib: place a thin wedge under the ipsilateral shoulder to slightly elevate the shoulder on the affected side — this opens the scalene triangle and reduces compression during imaging, providing a "decompressed" view for comparison with the neutral position view.

For tall patients whose axillary plexus falls below the inferior coil coverage: recentre inferiorly and repeat the coronal sequences with an extended inferior coverage slab, accepting reduced skull base coverage at the superior margin.

For the 3D CISS sequence at 3T when banding artefacts are visible: adjust the centre frequency (frequency offset) by ±100–200 Hz to shift the banding artefact away from the foraminal level where it is most diagnostically critical. This is a scanner-specific adjustment requiring advance familiarity with the 3D CISS/FIESTA-C banding behaviour on the local system.

8.3 Fast Salvage Protocol

Approximately 18 minutes — adequate for traumatic injury pre/postganglionic assessment and basic plexus characterisation. Does not provide axial T1 or post-contrast assessment.

8.4 Common Avoidable Errors

9. Quality Control Checklist

• C4–C5 foramina included in superior coverage (verify on coronal T1)
• Axillary level included in inferior coverage
• Coronal slab tilted to follow plexus obliquity (10–20° anterior-inferior tilt on sagittal localiser)
• Phase encoding R-L for coronal sequences
• Phase encoding A-P for axial sequences
• STIR fat suppression at cervicothoracic junction: acceptable / documented failure
• 3D CISS: foraminal level C5–T1 root sleeves visible bilaterally
• 3D CISS: banding artefact at foraminal level: absent / corrected / documented
• STIR acquired before gadolinium injection
• Denervation oedema assessed on STIR (paraspinal, periscapular, rotator cuff)
• Both clavicles visible on coronal images
• Correct laterality (L/R) labelled on axial sequences
• Post-contrast sequences acquired at appropriate timing if indicated

11. Evidence Gaps and Ongoing Debate

3D CISS vs CT myelography for root avulsion: 3D CISS/FIESTA-C has largely replaced CT myelography at expert centres for preganglionic injury assessment [3, 4], but formal prospective comparison demonstrating equivalent or superior diagnostic accuracy for pseudomeningocele detection across centres has not been systematically published. CT myelography remains the reference standard in centres without access to high-quality 3D isotropic MRI.

MR neurography standardisation: multiple 3D neurography sequences (3D STIR, 3D DWIBS, 3D MRN with fat suppression and diffusion-weighted nerve isolation) have been described for nerve delineation. No comparative prospective study has identified a single standardised neurography sequence that outperforms STIR + 3D CISS for brachial plexus clinical assessment.

ADC values in brachial plexus pathology: DWI-based ADC quantification for individual nerve trunks has been described in small series. Normal nerve ADC values, pathological thresholds for malignant infiltration, and radiation neuropathy vs recurrence are not standardised; the resolution required to measure ADC within a 3–5 mm nerve trunk exceeds what is achievable with clinical EPI-DWI at most centres.

Ultrafast MR neurography: compressed sensing and simultaneous multislice techniques may reduce the 3D CISS acquisition time to < 4 minutes at sub-millimetre isotropic resolution. Whether this enables full plexus assessment in a shortened overall protocol without diagnostic penalty has not been validated in clinical populations.

Post-radiation plexopathy vs recurrence: distinguishing post-radiation fibrosis (T2-dark, non-enhancing) from tumour recurrence (T2-bright, enhancing) remains imperfect, particularly for intermediate presentations. The reported sensitivity and specificity of MRI in this distinction (approximately 80–90% in expert series) has not been validated in multicentre prospective studies.

Related salivary gland master: Parotid Glands MRI — Generic Standard Protocol.