Neuromuscular ultrasound of the brachial plexus: know-how, know-when

Abstract

Neuromuscular ultrasound has become an integral diagnostic tool for brachial plexopathies, complementing electrodiagnostic studies and MRI. Despite the complex anatomy of the brachial plexus, nearly its entire postganglionic course is accessible by ultrasound, providing real-time, high-resolution structural information at the bedside. Characteristic sonographic findings have been described in traumatic plexopathies, neurogenic thoracic outlet syndrome, neuralgic amyotrophy, immune-mediated polyneuropathies, and neoplastic plexopathies. Evidence of moderate-to-high certainty now supports the use of ultrasound as a diagnostic adjunct in these conditions. Integration into clinical practice requires thorough knowledge of plexus anatomy and sonoanatomy, appropriate scanning techniques, and familiarity with condition-specific sonographic patterns. This review provides a practical framework for brachial plexus ultrasound, covering applied anatomy, systematic scanning approach, sonographic parameters and their interpretation, condition-specific findings, scanning protocols, and an evidence-based summary of its current indications.

Highlights

• •A standardized scanning technique for the brachial plexus is recommended.
• •Sonographic findings in brachial plexopathies impact diagnosis and management.
• •Evidence of moderate-to-high certainty supports use of brachial plexus ultrasound.

1. Introduction

Neuromuscular ultrasound has emerged as a valuable adjunct to electrodiagnostic studies in the workup of peripheral nerve and muscle disorders, including brachial plexopathies (van Alfen et al., 2018; van Alfen and Mah, 2018; Symanski et al., 2022; Bourke et al., 2024; Le Corroller, 2024; Yip et al., 2024). While electrophysiological studies provide indispensable information on the underlying pathology, severity, chronicity, and localization of brachial plexopathies, imaging adds the structural dimension necessary for a comprehensive assessment. MRI remains the gold standard for assessment of the preganglionic brachial plexus, but beyond the spinal canal ultrasound offers distinct advantages. Ultrasound is a bedside, affordable, and dynamic tool that provides high spatial resolution for imaging postganglionic brachial plexus. Characteristic structural findings in the plexus have been reported across the main categories of brachial plexopathies including, traumatic injuries, immune-mediated polyneuropathies, neuralgic amyotrophy, and neurogenic thoracic outlet syndrome (Arányi et al., 2015; Grimm et al., 2016; Goedee et al., 2017; Arányi et al., 2017; Chen et al., 2023; Feaster and Meiling, 2025). This practical review provides an in-depth resource for brachial plexus ultrasound covering applied anatomy, scanning techniques, sonographic parameters, condition-specific findings, and a summary of the evidence-based indications.

2. Applied anatomy of the brachial plexus

The brachial plexus is a complex neural network that runs along the neck down to the axilla and is responsible for innervating the entire upper limb. The plexus is formed by the ventral rami of the C5 to T1 cervical roots, with occasional contribution of C4 and T1 roots, known as a prefixed and postfixed plexus, respectively. The C5 and C6 roots unite to form the superior trunk, the C7 root forms the middle trunk, and the C8, T1 roots unite to form the inferior or lower trunk. Each trunk then divides into anterior and posterior divisions. The anterior divisions of the upper and middle trunks form the lateral cord, the anterior division of the lower trunk becomes the medial cord, and the three posterior divisions of the three trunks unite to form the posterior cord. Looking closely, the anatomy of the plexus is far more complex than the simple description of formation, because of the relation of the plexus to many other anatomic structures, including muscles, blood vessels, and bones, and the wide range of anatomic variations that can be encountered. A thorough knowledge of the plexus course in relation to the surrounding structures is a mandatory prerequisite for ultrasound scanning of the plexus.

The plexus can be divided into three parts: the supraclavicular part which comprises the roots, trunks, and divisions; the clavicular part running deep the clavicle; the infraclavicular part which comprises the cords; and the axillary part which represents the terminal branches.

The roots exit the cervical spine at the proximal part of the posterior neck triangle, known as the extraforaminal level, and they are situated between the anterior and the posterior tubercles of the corresponding transverse processes. The C5, C6, and C7 roots are found at their corresponding transverse process. However, the C8 root is situated below the C7 transverse process, and the T1 root originates from underneath the first rib (below the T1 costovertebral junction). The roots then course in the interscalene groove to form the trunks that run between the anterior scalene anteriorly and middle scalene posteriorly, deep to the sternocleidomastoid and lateral to the carotid sheath. Worth noting, the phrenic nerve runs on the anterior surface of the anterior scalene, and the dorsal scapular and long thoracic nerves traverse the middle scalene.

Further distal at the level of the supraclavicular fossa, each trunk divides into anterior and posterior divisions that descend lateral to the subclavian artery over the first rib. At the same level, the inferior belly of the omohyoid muscle crosses superficial to the plexus, with the suprascapular and long thoracic nerves running deep to it.

Just distal to the clavicle, the divisions form the cords and run lateral to the subclavian artery, deep to the pectoralis major and subclavius muscles. Then the cords start to separate and orient themselves around the axillary artery at the retropectoral groove deep to the pectoralis major and minor muscles.

Next, the plexus courses distally to the axilla, where it gives off its main terminal branches that have a relatively fixed orientation in relation to the axillary artery. The lateral cord gives the lateral root of the median nerve and musculocutaneous nerve, the medial cord gives the medial root of the median nerve, the ulnar nerve, the medial cutaneous nerve of the arm, and the medial cutaneous nerve of the forearm, while the posterior cord gives the axillary and radial nerves.

Regarding the fascicular topography, the ratio between neural and non-neural tissue and the number of fascicles vary at different levels (Moayeri et al., 2008; Orebaugh and Williams, 2009) For example, nerve roots lack the thick, hyperechoic connective tissue sheath found in peripheral nerves (Bonnel, 1984), while the cords have the largest number of fascicles (Johnson et al., 2010). Such arrangement reflects on the sonographic appearance of the plexus at different levels.

3. Brachial plexus scanning

3.1. Machine settings & image optimization

Typically, a high-frequency (> 12 MHz) linear ultrasound probe with a 4–5 cm footprint is used to scan the brachial plexus in adults but the frequency selection depends on the scanning plane, scanning level, and the depth of the target structures. Lower frequencies may be needed when scanning patients with a muscular or large neck, while ultra-high frequency probes – if available - allow better visualization of small nerves such as the phrenic, long thoracic, and dorsal scapular nerves. In neonates and children, a hockey stick probe with a small footprint may be more appropriate, especially in longitudinal scanning along the spine.

Currently, many ultrasound systems lack dedicated nerve presets, and therefore the musculoskeletal preset is often used. Because the musculoskeletal settings tend to “glue” soft tissue layers and use compounding techniques, turning off or down some of these image optimizers and decreasing the dynamic range (i.e., “choosing fewer shades of gray”), is recommended to achieve the maximal spatial resolution needed for nerve visualization.

To further optimize the image, basic machine knobs, including the depth, focus, gain, and time gain compensation, will need to be adjusted. The optimum depth should centralize the plexus on the screen, typically set between 1 and 4 cm for average-built adults and up to 6 cm for the tall and large-built people to visualize the deep interscalene C8 and T1 elements. A single focal point set at the level of the plexus optimizes the lateral resolution. The gain is best set so that there is a clear contrast between the plexus and the surrounding structures. The ‘Autogain’ knob in some machines can be helpful, but manual adjustment might be needed at different scanning levels. Time gain compensation knobs are usually kept in the middle (mid-range of brightness) or moved to the right or left to amplify ultrasound waves or eliminate noise in a specific layer. Re-adjustment of the frequency, depth, focal position, and gain is usually needed as one scans the plexus from one level to another.

Color Doppler is essential to identify the vascular landmarks, and to distinguish small perineural vessels from plexus fascicles. When using Doppler, pressure by the probe should be minimized. The Doppler box should be positioned at the top of the screen and then its size is adjusted to include the structure of interest. To set Doppler sensitivity, the color gain is increased until random noise appear, then is reduced until the noise disappears. Using low pulse repetition frequency (PRF), low wall filter, high frequency, and small color box are recommended to achieve maximum sensitivity.

3.2. Scanning principles

The general rules of peripheral nerve scanning apply to brachial plexus scanning. A generous amount of gel is used to ensure good contact between the probe and the skin and avoid having to push hard with the probe, especially when scanning the plexus at the supraclavicular fossa. Exerting the least pressure by the probe and keeping the probe orthogonal to the plexus (rather than to the skin) ensure accurate cross-sectional area measurements. The plexus should be scanned in short- and long-axis views, but the clavicle prohibits capturing the full long-axis view at the supraclavicular fossa. When scanning in short axis, the probe orientation mark can be directed so that either the right side of the patient or the medial structures are always displayed on the left side of the image. When scanning in the longitudinal view, the marker is typically directed proximally towards the patient's head.

3.3. Acoustic windows

The brachial plexus is accessible through three acoustic windows: the supraclavicular, infraclavicular, and axillary windows. Through the supraclavicular window, the cervical roots, trunks, and divisions can be visualized at three consecutive planes from proximal to distal: the extraforaminal level, the interscalene interval, and the supraclavicular fossa, respectively. The infraclavicular window includes the costoclavicular level immediately distal to the clavicle, and the retropectoral level, where the plexus runs under the pectoralis major and minor muscles. The axillary window is where the terminal branches of the plexus can be visualized at the axilla in relation to the axillary artery.

3.4. Patient position

While the supine position with the patient's head rotated and slightly extended is a standard view (Baute et al., 2018), it may not be the best for patients with limited neck rotation and elevated shoulder and may not be comfy to all sonographers. An alternative approach is to scan while patient sit on a chair facing the ultrasound screen. The examiner is typically positioned slightly besides and behind the patient, either standing or sitting on a stool, so they also face the screen and operate the machine with their free hand (Fig. 1). This setup allows gravity to pull the shoulder down which maximizes visualization, does not need neck rotation, and gives the patient the chance to see what's on the screen.

Fig. 1.

Patient's position for brachial plexus scanning.

3.5. How to identify the plexus in the ultrasound image?

Identification of the brachial plexus in an ultrasound image depends on knowledge of its echogenic properties and anatomical landmarks. The echogenic property of any tissue depends on how the tissue interacts with the ultrasound waves. The fascicles within the plexus appear hypoechoic in ultrasound because they allow the absorption of the ultrasound waves, while the perineurium and epineurium appear hyperechoic because they reflect the ultrasound waves. The sonographic appearance of the plexus differs from one level to another and depends on the fascicular arrangement and neural-to-non neural tissue ratio, as will be described in the following section (Sections 3.6.1, 3.6.2, 3.6.3). The anatomical landmarks also differ from one level to another. At the supraclavicular, costoclavicular, infraclavicular, and axillary levels, blood vessels (subclavian or axillary artery) are used as the landmarks. At the interscalene interval level, the anterior and middle scalene are the landmarks, while the shape and size of the tubercles of the transverse processes of the vertebrae are used as the landmarks to identify the levels of the cervical roots. Finally, differentiating the neural elements from small perineural blood vessels is easy by observing the arterial pulsation, venous compressibility, and Doppler signal.

3.6. Systematic scanning protocol

Diagnostic scanning of the brachial plexus can feel daunting for sonographers because of the intricate anatomy, the anatomic variations that are the rule rather than the exception, the relatively long course of the plexus, and the continuous change in the plexus course, which necessitates constant adjustment of the probe position and angle. For all these reasons it makes sense to have a fixed roadmap for the sonographer to rely on.

3.6.1. Supraclavicular window

For a basic interscalene and supraclavicular scanning protocol, we would recommend the following approach. With the patient positioned sitting in front of the ultrasound screen, start by positioning the probe in a transverse orientation over the trachea in the middle of the screen. Slide the probe cranially and caudally to find the thyroid gland and isthmus. The thyroid gland appears as a homogenous, smooth, relatively hyperechoic glandular tissue consisting of two lobes with the isthmus appearing as a bridge of glandular tissue connecting the left and right lobes (Supplementary Fig. 1). The thyroid gland isthmus typically corresponds to the C6-C7 level (Badshah et al., 2017). Lateral to the thyroid gland, the common carotid artery, the internal jugular vein can be seen enclosed within the carotid sheath with the vagus nerve wedged posteriorly between them (Supplementary Fig. 2). From this transverse position, rotate the probe out laterally around the neck while keeping it horizontal at the isthmus level. This will bring the scalene muscle region into view. Keep moving laterally until the common carotid artery is at the medial edge of the image. At this point, slowly incline the probe cranially to a 30° - 45° angle to bring the C5 – C6 – C7 elements into view in the interscalene groove, i.e., the fascial plane between the anterior and middle scalene muscles. The roots or the trunks at this level appear as hypoechoic, monofascicular structures, stacked over each other and arranged in a topographically predictable fashion, meaning the C5 elements will be most superior, the C6 elements just deep to them, and the C7, C8, and T1 elements deep to the C6 ones (Fig. 2). This view provides the home base for the scanning roadmap of the plexus. The next maneuver is to first bring the supraclavicular trunks/divisions in view, by moving the probe out towards the acromion while simultaneously tilting it progressively more upright, usually to an 60° - 80° angle, resembling a “slow ice cream scoop” motion, to end up with a view of the whole brachial plexus lying lateral to the subclavian artery over the first rib (Fig. 3). The middle scalene can be seen posterior to the plexus, while the anterior scalene is seen anterior to the subclavian artery, separating it from the subclavian vein. If plexus visibility is lost as one goes from the root level to the supraclavicular fossa, the probe can be directly placed at the root of the neck to first identify the subclavian artery, then look for multiple round or oval hypoechoic structures superolateral or lateral to the artery representing the plexus. Here, the neural elements are grouped together as a cluster, giving the appearance of a bunch of grapes (Fig. 3). This deep interscalene level represents the divisions of the plexus with the nerve elements anatomically organized. The most superficial elements represent the superior trunk/C5,6 roots, the elements just deep to them represent the middle trunk/C7 root, and the deepest elements wedged in the corner between the rib and the artery represent the lower trunk/C8, T 1 roots (Supplementary Fig. 3). The superior elements usually appear as three fascicles lying horizontally and represent the anterior division, the posterior division, and the suprascapular nerve (Supplementary Fig. 4). If one traces the superficial elements back and forth, the suprascapular nerve can be seen as the most lateral fascicle diverting from the superior neural elements running under the omohyoid muscle on its way to the supraspinous fossa (Supplementary Fig. 4).

Fig. 2.

Short-axis view of the cervical roots at the interscalene interval. AS = anterior scalene; MS = middle scalene.

Fig. 3.

Short-axis view of the brachial plexus at the supraclavicular fossa ‘deep interscalene level’. Arrow = brachial plexus, SA = subclavian artery.

From this supraclavicular view, the next step is to trace the individual nerve elements back to their respective root levels. To do this, center the element of interest in the middle of the ultrasound screen in a transverse view, and adjust the probe angle until the element appears as bright as it can be, indicating a perpendicular 90° scan angle, while sliding the probe cranially. The key to this continuous tracing technique is visual tracking of each group of the neural elements. Starting with the top components of the bunch of grapes; they can be seen separating from the cluster to appear as a distinct upper trunk at the interscalene level. On further proximal tracing, the C6 root will appear in view, followed by the C5 root further cranially. Similarly, visual tracking of the middle elements of the bunch of grapes ends in identification of the middle trunk at the interscalene level, then the C7 root – usually split in two distinct fascicles - will appear in view further proximally. Visual tracking of the deepest elements of the bunch of grapes may allow visualization of the lower trunk, followed by visualization of the C8 root and sometimes the T1 element on further proximal tracing. The C8 element emerges from deep in the interscalene region over the pleural dome; T1 is typically visualized only briefly as it curves behind the first rib in the supraclavicular region.

The identification of C7, C6, and C5 roots can be verified by looking at the size and shape of the anterior and posterior tubercles of the transverse process (Fig. 4). The tubercles appear as small hyperechoic curved lines with posterior acoustic shadowing. Of note, the optimal scan angle for the bones is usually different from the angle needed to visualize the roots. The transverse process of C7 lacks an anterior tubercle but has a prominent posterior tubercle that resembles the back of a lounge chair. Here, the vertebral artery that is commonly seen anterior to the C7 root, can be used as an additional landmark. The transverse process of C6 is characterized by a prominent anterior tubercle, known as “the carotid tubercle”, and a smaller posterior tubercle, creating a “thumbs-up”-like appearance. The C5 root is typically smaller in size than C6 root, and its transverse process has equally sized anterior and posterior tubercles creating a “molar tooth-like” appearance.

Fig. 4.

Short axis view of the C7, C6, and C5 roots at the extraforaminal level. The characteristics of the tubercles of the transverse processes are shown. AT = anterior tubercle; PT = posterior tubercle.

To capture the cervical roots in the long-axis view, the probe is centered on one of them and rotated 90 degrees from the short-axis view. Another way is to first capture the vertebral artery in the longitudinal view to see it interrupted at regular intervals by the acoustic shadows of the anterior tubercles (Bianchi and Martinoli, 2007). Sliding the probe slightly posterior will move the artery out of the view, and the roots appear as hypoechoic tubular structures exiting the neural foramina (Supplementary Fig. 5).

Several anatomic variants can be encountered during scanning the plexus along the supraclavicular window, including vascular structures traversing the plexus (Supplementary Figs. 6 and 7), anomalous medial insertion of the middle scalene muscle onto the first rib, an accessory small muscle intervening between the plexus and the subclavian artery called the scalenus minimus, cervical rib, and elevated first rib.

3.6.2. nfraclavicular window

With the shoulder abducted and externally rotated, the probe is placed medial to the coracoid process in a “lateral - sagittal” orientation so that the probe is transverse to the plexus course, not to the chest (Fig. 5). Increasing the depth and lowering the frequency are usually necessary at this level given the deep position of the plexus elements, that can be around 5–6 cm around the subclavian artery.

Fig. 5.

Short-axis view of the brachial plexus cords at the retropectoral level. AA = axillary artery; LC = lateral cord; MC = medial cord; PC = posterior cord. The image at the left bottom side shows the probe position.

The axillary artery and vein, and the pectoralis major and minor muscles are used as landmarks. The axillary vessels are first identified deep the pectoralis major and minor muscles. Next, the cords can be visualized around the artery. The lateral cord is the easiest to identify, superolateral to the artery; the medial cord lies medial to the artery or wedged between the artery and axillary vein; and the posterior cord lies posterior to the artery at the 6 o'clock position (Fig. 5). The cords may acquire the typical honeycomb appearance or may appear hypoechoic and monofascicular.

An alternative way is to track the cords from the costoclavicular level by placing the probe just distal to the clavicle looking for the axillary artery, that lies deep to the pectoralis major and subclavius muscle. The plexus can then be visualized as a cluster of neural elements lateral to the artery, similar to its appearance at the supraclavicular fossa (Fig. 6). At this level, the neural elements are topographically arranged in a different way than at the supraclavicular level. The most superficial and medial fascicles of the cluster represent the lateral cord, the inferior and medial fascicles represent the medial cord, which lies directly posterior to the lateral cord, while the most lateral fascicles represent the posterior cord (Sala-Blanch et al., 2016) (Supplementary Fig. 8). If the fascicular cluster at the costoclavicular level is traced distally, the cords will be seen separating from each other to orient around the axillary artery as described above.

Fig. 6.

Short-axis view of the brachial plexus at the costoclavicular level. Arrow = brachial plexus; AA = axillary artery; P.major = pectoralis major; Subclav = subclavius muscle.

3.6.3. Axillary window

With the patient in supine position and the shoulder abducted 90 degrees, the probe is placed in a transverse position to the axilla to identify the axillary artery. The terminal branches of the plexus can then be seen oriented around the artery. All branches at this plane are multi-fascicular, oval or rounded in shape with hyperechoic epineurium. The median nerve is typically lateral/superolateral, the ulnar nerve medial, and the radial nerve posterior to the artery (Fig. 7), while the musculocutaneous nerve lies lateral to the axillary artery and the median nerve in the fascial plane between the biceps and the coracobrachialis, though considerable variation exists (Retzl et al., 2001; Christophe et al., 2009). Tracing the nerves distally along the arm confirms their respective identity.

Fig. 7.

Short-axis view of the terminal branches of the plexus at the axilla. AA = axillary artery; AV = axillary vein; MN = median nerve; RN = radial nerve; UN = ulnar nerve; V = vein.

Tracing the plexus as a continuous structure from the infraclavicular to the axillary level or vice versa is typically difficult because of its deep position in the caudal infraclavicular region. However, it is possible for children and thin adults.

4. Brachial plexus sonographic parameters

The sonographic evaluation of the plexus includes assessment of plexus size, echotexture, and vascularity.

4.1. Size

Nerve size measurements of the brachial plexus are needed for the technique to be maximally sensitive to pathology. The international consensus is to use the cross-sectional area for this purpose (Tawfik et al., 2023). Similar to peripheral nerves, the cross-sectional of the plexus is best measured using the trace function of the ultrasound machine by following the neural elements without inclusion of the hyperechoic outer border. At the supraclavicular fossa and costoclavicular level, the cross-sectional area is measured by tracing the whole cluster (Supplementary Fig. 9 A), while the roots, trunks, and cords cross-sectional areas are measured as separate neural elements (Fig. 9B). The diameter of roots and trunks is measured using the distance function by extending the marks from just inside the hyperechoic rim.

Fig. 9.

Sonographic appearance of neuroma in-continuity. Long-axis view of C6 showing hypoechoic fusiform swelling in continuity with the root.

Two different flavors of nerve reference values exist (Goedee and van Alfen, 2023); population and disease-specific cut-offs. The population cut-off values (Grusauskiene et al., 2024; Wang and Wang, 2020; Cartwright et al., 2013; Pillen et al., 2015) are universally applicable to any patient from that population background and are non-specific indicators of pathology. In contrast, disease-specific cut-off values such as those created from patients with carpal tunnel syndrome or chronic inflammatory demyelinating polyneuropathy using receiver-operating characteristic curves, have been specifically chosen to reflect a known sensitivity and specificity, and should be used only in the same clinical context. Table 1 summarizes cross-sectional reference values at different anatomical levels extracted from several studies. However, it is generally recommended that each EMG lab or country establishes its own reference values.

Table 1.

Reference values of the cross-sectional areas of the brachial plexus elements.

Neural element	n	CSA (mm2)	Reference study
C5 root	95 volunteers	5.66 ± 1.02 a	Won et al., 2012
	43 volunteers	9.0 ± 2.3 a	van Rosmalen et al., 2019
	911 nerves	5.62 b	Fisse et al., 2021
	120 nerves	4.7/7.3 c	Takenaka et al., 2026
C6 root	95 volunteers	8.98 ± 1.65 a	Won et al., 2012
	44 volunteers	13.4 ± 3.6 a	van Rosmalen et al., 2019
	909 nerves	8.8 b	Fisse et al., 2021
	120 nerves	8.5/12.3 c	Takenaka et al., 2026
C7 root	95 volunteers	10.43 ± 1.86 a	Won et al., 2012
	42 volunteers	13.0 ± 4.1 a	van Rosmalen et al., 2019
	909 nerves	9.5 b	Fisse et al., 2021
	120 nerves	9.5/13.9 c	Takenaka et al., 2026
C8 root	120 nerves	8.7/12.5 c	Takenaka et al., 2026
T1 root *	120 nerves	5.2/7.8 c	Takenaka et al., 2026
Interscalene plexus #	45 volunteers	37.6 ± 13.2 a	van Rosmalen et al., 2019
Upper trunk	95 volunteers	16.7 ± 2.88 a	Won et al., 2012
	120 nerves	19.9/26.3 c	Takenaka et al., 2026
Middle trunk	95 volunteers	14.01 ± 2.70 a	Won et al., 2012
	120 nerves	13.1/17.9 c	Takenaka et al., 2026
Lower trunk	95 volunteers	13.75 ± 2.57 a	Won et al., 2012
Neural elements at the supraclavicular fossa #	41 volunteers	79.0 ± 20.1 a	van Rosmalen et al., 2019
	120 nerves	73.7/98.1 c	Takenaka et al., 2026

CSA = cross-sectional area; n = number; * T1 root measured at the supraclavicular region; # Neural elements measured collectively.

^aValues are mean ± SD, ^b values are mean, ^c values are mean/upper limit.

Because the plexus continuously branch out and come together as they travel from the root foramina out to the limb, as known from the landmark work of Kerr in 1918 (Tritsch et al., 2024), a practical approach to assess the pathology is to measure the extraforaminal root cross-sectional area and apply two simple rules: 1) the cross-sectional area of an interscalene or supraclavicular element should not exceed the cross-sectional area of the root it came from, and 2) any nerve element should not have a focal or segmental enlargement of >50% of its size (or a ratio exceeding 1.5) when compared to a more proximal segment.

Most disorders, e.g., inflammatory or hereditary neuropathies, tumors, and trauma, result in focal, multifocal or diffuse nerve enlargement. It is important to realize that nerve enlargement is non-specific, and very different disorders may lead to similar sonographic size abnormalities, just as low motor and sensory nerve conduction amplitudes are non-specific indicators of axon loss. This means that all nerve ultrasound findings should only be interpreted in the clinical context of that particular patient.

There is one specific type of nerve pathology that leads to small nerves throughout the body, and that is sensory neuronopathy (Pelosi and van Alfen, 2023), which can also be picked up in the brachial plexus area. As can be theorized from nerve histology, axons make up about 40–60% of the nerve that's visible on the ultrasound screen, with connective tissue elements making up the rest of the nerve volume. Since 90% + of all axons in any nerve are afferent (Gesslbauer et al., 2017), losing these cells in dorsal root ganglionopathy will mean that the nerve will decrease around 45% in cross-sectional area.

4.2. Echotexture

The internal echotexture, as well as the echotexture and thickness of the outer connective tissue layer are typically assessed. Most plexus pathologies result in fascicular and axonal edema reflected as hypochogenicity of the neural elements. Normally, the roots/trunks appear as hypoechoic monofascicular structures; therefore, pathological hypochogenicity at these levels should be interpreted with caution and should not be used as a sole criterion of abnormality. Hyperechogenicity of the neural elements can be seen in cases of intraneural fibrosis as in radiation-induced plexopathy. Fascicular enlargement is characteristic of immune-mediated polyneuropathies, while increased thickness of the epineurium suggests chronicity.

4.3. Vascularity

The brachial plexus is highly vascularized, receiving blood supply from the subclavian and axillary arteries. The nerves in general, including the plexus are surrounded by a network of perineural arteries that run longitudinally along the epineurium and perineurium and intraneural tiny penetrating branches that extend inward among the fascicles (Valle and Zamorani, 2007; Gilcrease-Garcia et al., 2020). Normally, the brachial plexus elements show no intraneural Doppler signal. Increased flow may be seen in tumors, intraneural vascular malformations, and active immune-mediated neuropathies (Graif et al., 2004; Carandang et al., 2017).

5. Condition-specific findings and scanning protocols

5.1. Traumatic brachial plexopathies

Traumatic injury of the brachial plexus can be a devastating event leading to prolonged physical disability. The type and timing of the intervention can determine the prognosis. Thus, a thorough assessment of the functional and structural integrity of the plexus is needed to decide the optimal management of the injury. Nerve conduction and electromyography studies can help differentiate pre-ganglionic from post-ganglionic lesions, determine the underlying pathology, severity, and chronicity of the injury. However, electrodiagnosis is non-informative in the first two weeks after injury before Wallerian degeneration has occurred, sometimes fails to localize the level of the lesion, and cannot reveal structural abnormalities that might interfere with further reinnervation even in cases of partial axonal lesions. Ultrasound represents an attractive bedside tool to assess the structural integrity of the plexus at any timepoint following the injury, and can even identify hidden injuries that are missed by EMG (Meiling et al., 2026).

The spectrum of ultrasound findings varies according to the type and severity of the injury. Sunderland grade I or II (Neuropraxia/mild axonotmesis) injury results in either a normal appearance and size of the neural elements or a mild nerve swelling with preserved fascicular echotexture. Sunderland grade III manifests as a neural swelling with altered echotexture (Fig. 8), that is usually hypoechoic in the acute phase. A neuroma in-continuity denotes Sunderland grade IV and appears as an ovoid or fusiform, hypoechoic, tumor-like lesion with a disturbed fascicular architecture on long axis views in continuity with the plexus (Gruber et al., 2007) (Fig. 9). In contrast, neural element rupture, complete transection, and root avulsion denote high-grade injury, i.e., a Sunderland grade V (Wijntjes et al., 2020; Goyal et al., 2021). Rupture without epineural avulsion appears as an “empty neural bed” with no internal fascicles (Fig. 10). In complete root injury, an abrupt swollen end of the plexus element with or without neural gap is seen (Fig. 11). Stump neuromas may also form at the transected ends in the subacute or chronic stage (Gu et al., 2022). Partial disruption of the epineurium suggests incomplete transection; this will still need surgical repair. Pre-ganglionic rootlet avulsion is not directly visualized via ultrasound, but pseudomeningocele may be seen at the interscalene level. If this type of injury is clinically suspected, for example in patients with concomitant cervical spine injury and complete paralysis and anesthesia of one or more levels, an MRI or CT myelography should be added to the workup to assess the intraspinal rootlets. Pseudomeningocele appears as a mass showing accumulation of cerebrospinal fluid with absence of neural structures inside (Gu et al., 2022). Scar tissue compromising the plexus appears as an irregular hypoechoic tissue encasing or overlying the plexus element (Fig. 12).

Fig. 8.

Sonographic appearance of plexus traction injury. Long-axis view of the C6 root (arrows) showing swelling of the root.

Fig. 10.

Sonographic appearance of neural element rupture. Short-axis view of the trunks of the brachial plexus at the interscalene groove showing empty beds of the upper and middle trunks. LT = lower trunk; MT = middle trunk; UT = upper trunk.

Fig. 11.

Sonographic appearance of complete plexus transection. Long-axis view of the upper trunk (white arrows) and middle trunk (green arrows) showing discontinuity of the two trunks. MT = middle trunk; UT = upper trunk. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 12.

Sonographic appearance of cord pinching. Long-axis view of the lateral cord at the retropectoral space showing indentation of the cord (solid arrows) by longitudinal scar tissue (dashed arrow).

In neonatal or obstetric brachial plexus injuries, the trunks or roots may appear swollen with or without distorted echogenicity in partial injuries (Tawfik et al., 2020), but neuroma in-continuity commonly forms and typically involves the upper trunk (Somashekar et al., 2016). Of note, ultrasound detection sensitivity is higher for upper and middle trunk neuroma (84%) than for lower trunk neuroma (68%) (Somashekar et al., 2016). Worth mentioning, the tubercles of the transverse processes are cartilaginous in children and thus appear hypoechoic rather than hyperechoic (Somashekar et al., 2016). In addition, neonates have an extra tissue layer in the neck just deep to the sternocleidomastoid muscle that is brown fat (Pillen et al., 2015) which needs to be considered when interpreting the regional anatomy.

The scanning protocol for traumatic brachial plexopathy is quite similar to the general scanning protocol of traumatic nerve injuries (Tawfik et al., 2023), but few protocol steps can be challenging in traumatic plexopathies such as acquiring the longitudinal view and clear visualization of the infraclavicular portion of the plexus and C8, T1 roots.

Using a strict sterile approach is not needed except when scanning wound areas. It is always better to start scanning in short-axis view at a level remote from the expected injury site, then move towards the injury level to capture the pathology. The first step is to assess the continuity of the neural elements. This is best achieved by back-and-forth tracing of the plexus and observing if there is a sudden loss of the neural elements visibility, which may suggest rupture or complete plexus transection. It should be noted that lost visibility of the neural elements may be due to a masking effect of a scar tissue. Therefore, the probe must be rotated 90 degrees to confirm the discontinuity in the longitudinal view when possible. If transection is confirmed and a nerve gap is observed, the levels of the proximal and the distal neural stumps are determined, and the length of the nerve gap is measured in the longitudinal view. The proximal and distal stumps should also be checked for the presence of a stump neuroma. Of note, the stumps often do not align longitudinally anymore following transection but may have shifted slightly lateral or medial with respect to each other. Measuring the neuroma size in the short-axis view, and its length in the long-axis view is important for surgical planning.

If the plexus is found to be in continuity, the next step is to check for partial disruption of the outer epineurium, which could suggest partial transection; neuroma in-continuity; or encasement of the neural elements by scar/fibrous tissue. Gently exerting slight pressure with the probe (i.e., “sonopalpation”) on the plexus at the pathology site may trigger sensory symptoms and help in detecting clinically relevant pathologies. Providing external or internal anatomic landmarks to describe the level of the abnormal sonographic findings is highly recommended for pre-surgery planning. Scanning the surrounding structures is also important to screen for any associated abnormalities in the surrounding tissues, such as foreign bodies including glass shards or metal fragments. Lastly, asseement of the echotexture of different muscles innervated by the different portions of the plexus can indirectly help in localizing the lesion if needle electromyography has not been performed or does not provide useful information.

5.2. Neurogenic thoracic outlet syndrome

The thoracic outlet syndromes are a group of disorders all caused by compression of nerve and/or blood vessels as they travel through the thoracic outlet space, spanning the interscalene triangle in the neck, the supraclavicular areas, and the subpectoral infraclavicular region of the chest wall (Attaar and Pascarella, 2025). The most common form of thoracic outlet syndrome is associated with neurologic symptoms of paresthesia, numbness, and weakness, and therefore called neurogenic thoracic outlet syndrome. The diagnosis of neurogenic thoracic outlet syndrome depends on clinical examination, including special provocative tests, electrodiagnostic studies, and imaging. Nerve conduction and electromyography can confirm diagnosis by documenting a characteristic pattern of median nerve motor affection, ulnar nerve sensory affection, and an abnormal medial antebrachial cutaneous nerve sensory study, determine the severity, chronicity, and underlying pathophysiology, can differentiate between thoracic outlet syndrome & its mimics like C8,T1 radiculopathy, and can rule out peripheral local entrapment & double crush lesion. However, the classic electrodiagnostic pattern, once thought pathognomonic for true neurogenic thoracic outlet syndrome, probably only detects about half of these patients (Mul et al., 2022). Therefore, other diagnostic tests are needed. Although MRI is widely used, ultrasound can provide valuable information and has been listed as one of the useful diagnostic tools in neurogenic thoracic outlet syndrome by the Section of Peripheral Nerve Surgery of the European Association of Neurosurgical Societies and International Neurogenic Thoracic Outlet Syndrome Hand Surgery Workgroup (Dengler et al., 2022; Chim and Hagan, 2024). Ultrasound can localize the level of compression and detect anatomic variants that predispose to neurogenic thoracic outlet.

The most commonly encountered muscle variant is anomalous medial insertion of the middle scalene in which the middle scalene muscle protrudes medially onto the first rib elevating the plexus and the subclavian artery (Makhoul and Machleder, 1992; Martínez-Sanz et al., 2025) (Fig. 13), and can cause nerve elements to deviate off course and be entrapped A hallmark sonographic sign associated with this variant has been dubbed the “wedge sickle sign” (Arányi et al., 2016) with the wedge being the protruding muscle and/or fibrous band and the sickle being the kinked or curved appearance of the nerve element(s) around it (Fig. 14). Pressure by the probe on the lower trunk at the indentation level may provoke neuropathic symptoms along the arm known as “ultrasound Tinel's sign” (Arányi et al., 2016). Scalenus minimus, which is an accessory muscle, can occasionally be encountered at the supraclavicular fossa intervening between the plexus and the subclavian artery.

Fig. 13.

Anomalous medial insertion of the middle scalene muscle. Short-axis view of the brachial plexus at the supraclavicular fossa showing anomalous medial insertion of the middle scalene (arrow) onto the first rib. MS = middle scalene, SA = subclavian artery.

Fig. 14.

Wedge sickle sign in neurogenic thoracic outlet syndrome. Short-axis view of the lower trunk just cranial to the supraclavicular fossa showing the fibromuscular band of the middle scalene (arrow) indenting and compressing the lower trunk (dashed arrow).

Ultrasound may also reveal bony anatomic variants, including cervical rib and elevated first rib, that can cause lower trunk compression (Chen et al., 2023; Dollinger et al., 2022). A cervical rib is visualized with ultrasound as an elongated posterior element of the C7 transverse process or as a bony, curved structure with posterior acoustic shadow encroaching on the middle scalene at the interscalene interval level (Fig. 15 A) or articulating with the first rib at the supraclavicular fossa, either lying posteromedial to the plexus or even pushing it up and narrowing the space available for the plexus (Fig. 15 B). The medial end of the first rib may also indent and compress the lower trunk/C8, T1 roots, which is termed the “bony wedge sickle sign” (Fig. 16).

Fig. 15.

Sonographic appearance of cervical rib. Short-axis view of the brachial plexus (dashed arrow) at the (A) interscalene interval, and (B) supraclavicular fossa showing the cervical rib (solid arrow). AS = anterior scalene; MS = middle scalene; SA = subclavian artery.

Fig. 16.

Bony Wedge Sickle sign. Short-axis view of the brachial plexus at the supraclavicular fossa showing indentation of the lower trunk (arrow) by the medial end of the first rib.

It should be noted that the described anatomic variants can be a risk factor for neurogenic thoracic outlet syndrome but they should not be interpreted as pathological per se as they can be encountered in healthy individuals without any symptoms. Therefore, correlation with the clinical and electrodiagnostic findings is crucial to avoid overdiagnosis.

The spectrum of sonographic findings in neurogenic thoracic outlet syndrome also includes increased cross-sectional area of brachial plexus cords on the affected side compared to the healthy side, either at the costoclavicular or retropectoral levels; and pectoralis minor muscle hypertrophy, or posterior indentation during progressive shoulder abduction, causing narrowing of the retropectoral space (Ahmed et al., 2022).

To scan for a true neurogenic thoracic outlet syndrome, starting from the basic protocoloutlined in Section 3.6., the approach would be to identify the oblique line of interscalene groove elements and scan out to the transition of the neck into the shoulder area, where all elements are lined up, and the C8 element can be seen appearing over the pleural dome. Color doppler is advised to distinguish this C8 element from the vertebral artery which can also be visible in this region. Next, the main step is to follow the course of the C8 element in its trajectory out to the supraclavicular region. Scan slowly, keep going back and forth, and look closely at whether the nerve element at some point needs to deviate medially off its straight course in the interscalene groove and out to the supraclavicular view. Deviation may be seen without nerve swelling, and also in asymptomatic people as an incidental finding, so this finding needs to be correlated with the clinical symptoms outlined above. Often, an elongated posterior element of the C7 transverse process (i.e., a cervical rib) is present, which can already be seen on coronal MRI images as a clue to look for a wedge sickle sign.

5.3. Neuralgic amyotrophy

Neuralgic amyotrophy, also known as Parsonage-Turner syndrome, is an acute-onset inflammatory multifocal peripheral neuropathy that most often affects the upper extremities (Gabet et al., 2025). Formerly known as “brachial plexus neuritis”, it has now become clear that the affected nerve segments are typically not localized to the brachial plexus itself, but rather affect the proximal main end branches (Sneag et al., 2018) including the long thoracic and suprascapular nerves in the scalene region.

Neuralgic amyotrophy has some unique imaging features that can also be seen on ultrasound as early as 12 h from symptom onset (Ripellino et al., 2022). This diagnostic imaging has the potential to significantly enhance patient care, as almost all neuralgic amyotrophy patients presenting acutely receive an incorrect diagnosis initially (Campbell et al., 2025). Typical ultrasound findings include nerve swelling, with a ratio of >1.55 compared to the contralateral side having a 95% sensitivity and 87% specificity for neuralgic amyotrophy (Yu et al., 2023), pathognomonic “hourglass” nerve constrictions that may warrant surgical intervention if no spontaneous recovery occurs over 4–6 months (Gstoettner et al., 2020; Krishnan et al., 2021), and intraneural fascicular rotation by >180 degrees called nerve torsion or fascicular entwinement (Fig. 17), that have not been found in any other nerve disorder so far.

Fig. 17.

Pathological sonographic findings in neuralgic amyotrophy. Long-axis view of the phrenic nerve at the neck showing torsion of the nerve fascicles (dashed arrows) and hourglass constriction (solid arrows).

The scanning protocol based on expert consensus should focus on the symptomatic nerves (Cignetti et al., 2023).

To pick up fascicular rotation and hourglass constrictions with ultrasound, a very slow scan speed is advised, with meticulous visual tracking of different nerve fascicles, their size, and position relative to each other. Fascicular rotation will cause the fascicles to appear to tumble clockwise or counterclockwise around each other when scanning across a nerve segment. As some fascicular reorientation is normal along the course of a nerve (Sunderland, 1990), only rotations involving >180 degrees should be considered potentially abnormal. Hourglass constriction is best captured in the long-axis view but is suspected in the transverse view when fascicular swelling is observed, followed by sudden narrowing or even disappearance of the fascicles in the nerve, only to reappear after few millimeters. In the longitudinal view, the constriction appears as an abrupt decrease in nerve diameter, which could be incomplete with preservation of neural continuity, or complete with disturbed fascicular continuity and interfascicular hyperechoic cleavage line (Fig. 17).

5.4. Brachial plexus findings in immune-mediated polyneuropathies

Immune-mediated neuropathies are a group of disorders in which the peripheral nerves are damaged by an abnormal autoimmune response. The most important disorders are Guillain-Barré syndrome, chronic inflammatory demyelinating polyradiculoneuropathy, and multifocal motor neuropathy (Willison et al., 2016). The diagnosis of immune-mediated polyneuropathies depends on the clinical history and examination, electrodiagnostic studies, and laboratory investigations or cerebrospinal fluid analysis. In classical types, the electrodiagnostic criteria requires unequivocal signs of demyelination including focal or multifocal conduction block. However, the conduction block may be in very proximal segments, and thus may not become apparent during nerve conduction studies. In such cases, imaging, with either ultrasound or MRI, can help reach the diagnosis.

In chronic inflammatory demyelinating polyneuropathy, the most consistent feature is enlargement of the brachial plexus trunks and roots (especially the C5–C7 roots) as the result of segmental demyelination and remyelination with or without onion-bulb formation (Vallat and Mathis, 2024). The enlargement is often bilateral and symmetric but sometimes asymmetric (Pitarokoili et al., 2023) (Fig. 18) with hypoechoic and enlarged fascicles, reflecting edema and inflammation. In more severe or chronic disease, the normal fascicular pattern may become blurred or lost (Oudeman et al., 2020).

Fig. 18.

Plexus sonographic findings in chronic inflammatory demyelinating polyneuropathy. Short-axis view of the enlarged trunks (A, C) and roots (B, D) of the brachial plexus.

In Guillain-Barré syndrome, the sonographic pattern is similar to chronic inflammatory demyelinating polyneuropathy (Fig. 19), but blurred fascicular borders can be seen early in the disease course. The sonographic changes develop rapidly, and may regress with recovery (Kerasnoudis et al., 2013).

Fig. 19.

Plexus sonographic findings in Guillain-Barré Syndrome. Short-axis view of the enlarged trunks (A, C) and roots (B, D) of the brachial plexus.

In multifocal motor neuropathy, the pattern is also quite similar to the chronic inflammatory demyelinating polyneuropathy pattern, but the neural elements' enlargement is typically focal and patchy or multifocal, less pronounced than in chronic inflammatory demyelinating polyneuropathies (Kerasnoudis et al., 2014) (Fig. 20), and often correlating with clinical weakness and conduction block sites (Breiner et al., 2019; Oudeman et al., 2020). The fascicular architecture is better preserved than in chronic inflammatory demyelinating polyneuropathy (Niu et al., 2025). The possible underlying pathological changes for the nerve enlargement include nerve fiber degeneration with significant reduction of large, myelinated fibers, areas of remyelination, clusters of regenerating axons, but onion bulb formation is usually not evident as in chronic inflammatory demyelinating polyneuropathy (Van Asseldonk et al., 2005; Claytor et al., 2025).

Fig. 20.

Plexus sonographic findings in multifocal motor neuropathy. Short-axis view of the enlarged trunks (A) and roots (B) of the brachial plexus.

Color Doppler may show increased blood flow within the enlarged plexus, especially in active or early stages, more commonly seen in chronic inflammatory demyelinating polyneuropathy and Guillain-Barré than multifocal motor neuropathy (Carandang et al., 2017).

Two scanning approaches can be followed in immune-mediated polyneuropathies. A validated short protocol (Herraets et al., 2020) involves measuring the cross-sectional area of the median nerve at the forearm and arm, and the C5 root. Unilateral or bilateral cross-sectional area enlargement of the median nerve at the forearm >10 mm², and/or median nerve at the arm >13 mm², and/or C5 root >8 mm² in a patient clinically suspected of chronic inflammatory demyelinating polyneuropathy supports the diagnosis with 85% sensitivity and 73% specificity (Herraets et al., 2020).

The second approach is a consensus-based extended scanning protocol (Tawfik et al., 2023) involving scanning the brachial plexus and at least two nerves in a limb along their courses with a focus on the most clinically involved nerves. Next, the cross-sectional area of the plexus and peripheral nerves is measured at pre-determined levels to determine the pattern of nerve enlargement (diffuse, focal, or multifocal, with analysis of the nerve echotexture, and vascularity.

5.5. Neoplastic brachial plexopathies

Neoplastic brachial plexopathies include primary neurogenic tumors, metastatic spread from other malignancies, invasion from nearby tumors, and radiation-induced plexopathies.

Primary brachial plexus tumors are rare, accounting for less than 5% of all tumors involving peripheral nerves (Huang et al., 2004), and most of them, usually schwannoma or neurofibromas, are benign in nature(Das et al., 2007). Malignant peripheral nerve sheath tumors are much less common but are important clinically due to their aggressive nature (Jia et al., 2016).

Ultrasound imaging has been shown to be a valuable, non-invasive diagnostic tool in the evaluation of brachial plexus tumors too. Early and accurate diagnosis of such masses is critical for appropriate management and optimal patient outcomes. Traditional imaging modalities such as magnetic resonance imaging and computed tomography have long been used in this setting; however, ultrasound offers unique advantages, including real-time assessment, detailed images of superficial structures with a 10 times higher resolution than a routine MRI, and the ability to guide directly diagnostic and therapeutic interventions.

With ultrasound, brachial plexus tumors are visualized as masses along the anatomical route of the brachial plexus element(s), often in the supraclavicular or infraclavicular regions. They usually appear as well-defined, fusiform, round or oval masses (Graif et al., 2004). Benign tumors tend to have smooth, well-circumscribed margins (Fig. 21), while malignant lesions often show irregular borders that do not respect the normal tissue boundaries (Graif et al., 2004) Most brachial plexus tumors are hypoechoic or show heterogeneous echotexture. A “target sign”, i.e. a central hyperechogenicity with peripheral hypochogenicity, may be seen in some cases (Fig. 21). Schwannoma is usually seen in continuity with a nerve fascicle, often with the nerve entering and exiting the lesion in a long axis view, with the lesion eccentric in the epineural sheath, displacing other fascicles. In contrast, neurofibroma often involves and expands multiple fascicles or the whole nerve, with the nerve fibers running through the tumor (i.e., central involvement). Color doppler imaging can show intraneural vascularity. Benign lesions tend to have low or peripheral blood flow, while malignant tumors may demonstrate increased, chaotic vascularity. Worth noting, ultrasound can help distinguish between solid and cystic components and may identify associated edema or invasion of surrounding tissues in more aggressive tumors. However, it should be noted that no single ultrasound feature is diagnostic of malignant peripheral nerve sheath tumors, and a definitive diagnosis requires biopsy and histopathology. MRI is also usually recommended for further characterization of the anatomic tumor extent (Siqueira et al., 2009).

Fig. 21.

Plexus Tumor. Short-axis view of the C7 root showing neurofibroma (arrow). Note the “target sign” in the center of the mass. CCA = common carotid artery; SCM = sternocleidomastoid.

Radiation-induced brachial plexopathy, a late effect of neck, chest, and axillary irradiation, develops as a result of intraneural and perineural fibrotic changes (Warade et al., 2019) that are caused by the disrupted tissue DNA repair induced by radiation. The critical point in management is to differentiate between radiation-induced plexopathy and tumor recurrence. This is typically achieved through an extensive workup including MRI, computed tomography scan, and positron emission tomography scan. However, ultrasound can be used as a screening test. In radiation-induced plexopathy, ultrasound shows hyperechoic thickening - fibrosis - of the neural elements of the plexus (Lapegue et al., 2014) (Fig. 22).

Fig. 22.

Radiation-induced plexopathy. Short axis view of the brachial plexus at the supraclavicular fossa showing hyperechoic thickening of the plexus and lost fascicular pattern. SA = subclavian artery.

In suspected brachial plexus tumors, scanning the entire course of the plexus is advised with back-and-forth tracing in the short-axis view, looking for an abrupt swelling of the neural elements. The next step is to rotate the probe in the long axis to confirm continuity of the mass with the neural element. The shape, echotexture, margins, and vascularity of the tumor are assessed to determine if the sonographic pattern is consistent with benign, or malignant nerve sheath tumor. In benign tumors, looking for the target sign and whether the mass is eccentric or central can help to tentatively discriminate schwannoma from neurofibroma. The scanning protocol should also include screening the surrounding structures for compression or invasion effect by the mass, and assessment of the supraclavicular, infraclavicular, and axillary lymph nodes.

6. Evidence-based indications of diagnostic ultrasound for brachial plexus pathologies

The indications of any diagnostic tool should be based on the evidence. Therefore, in this section we summarize condition-specific certainty of evidence for the utility of brachial plexus ultrasound. As this review is a structured narrative synthesis rather than a formal systematic review or guideline, a full GRADE process was not applied. Instead, certainty of evidence was assessed using a qualitative, GRADE approach. The evidence levels (low, moderate, high) were assigned based on overall appraisal of study design, risk of bias, consistency, directness to brachial plexus pathology, and outcome precision. This structured approach supports transparent interpretation of the role of neuromuscular ultrasound across different brachial plexopathy phenotypes and clinical contexts.

The details of the literature research methodology, the characteristics of the included studies, and the summary of the evidence are provided as supplementary material (“Literature review” file), while a summary of indications is listed in Table 2.

Table 2.

Evidence-based indications of brachial plexus ultrasound

1. Traumatic brachial plexus injury when there is a need to localize postganglionic structural lesions and especially when preoperative anatomical mapping is required.

2. Neurogenic thoracic outlet syndrome when the goal is to identify lower trunk compression and structural variants, particularly using the wedge-sickle sign, and when combined nerve + vascular ultrasound is feasible.

3. Clinically suspected neuralgic amyotrophy to provide structural confirmation and localization of affected nerves, particularly to detect hourglass constrictions, or torsion.

4. Suspected chronic inflammatory demyelinating polyneuropathy and multifocal motor neuropathy when electrophysiology is borderline or does not meet strict demyelination criteria.

5. Suspected early Guillain-Barré syndrome, particularly when nerve conduction studies are equivocal, delayed, or incomplete, to detect proximal and diffuse nerve enlargement (roots, vagus nerve, limb nerves).

7. Limitations of brachial plexus ultrasound

The main limitation of ultrasound of the brachial plexus is its inability to visualize the preganglionic part of the plexus. Visualizing the C8, T1 roots, and the retropectoral cords can also be challenging, particularly in large-built patients. In these situations, MRI – or ideally MR neurography - is recommended. Realizing that ultrasound and MRI are complementary rather than competing, allows the appropriate selection of both tools. Ultrasound serves well as a first-line imaging tool for postganglionic plexus and small neck nerves' pathologies given its accessibility and high spatial resolution, while MRI is better reserved for ambiguous cases, complex anatomy, and preganglionic lesions (Pacella et al., 2025).

8. How to integrate brachial plexus ultrasound in clinical practice?

Efficient integration of brachial plexus ultrasound requires solid knowledge of anatomy, cross-sectional anatomy, variants, structured supervised hands-on training, and clinical experience in correlating the sonographic findings with clinical and electrodiagnostic data.

A high-quality machine with a linear wide-band probe and frequency > 12 MHz, ideally >18 MHz is recommended. The brachial plexus is part of the curriculum of advanced-level neuromuscular ultrasound training in the consensus-based training guidelines (Tawfik et al., 2019), but basic scanning of the roots and trunks is frequently addressed in the basic- and intermediate-level training courses, given its importance in the diagnosis of immune-mediated polyneuropathies. Immediate practice of scanning on oneself, a colleague, or a healthy volunteer is advised after attending workshops to retain the skill. The sonographer should not start scanning patients and writing a report unless feeling confident of the skill. The learning curve varies; at minimum, six supervised examinations are needed to reliably identify structures at the axillary level (Helayel et al., 2009), with longer training expected for the more complex supraclavicular and infraclavicular regions. Mentorship, continued education, and adherence to standardized scanning protocols are advised to maintain competence.

In daily clinical practice, the ultrasound study is planned based on the referral letter and the clinical question. However, neurological examination to put down the differential diagnosis is mandatory to tailor the exam. Typically, nerve conduction studies and EMG are conducted prior to the ultrasound study, but ‘Ultrasound-first’ approach can be adopted in some cases like in traumatic plexus injuries to assess continuity of the plexus elements and in neuralgic amyotrophy to check for torsion or hour-glass constriction to decide the need for surgical intervention. The ‘Ultrasound-First’ approach is also suitable in children, or if the patient cannot tolerate the electrodiagnostic studies. Suggested flow charts illustrating how brachial plexus ultrasound integrates with electrodiagnostic studies and MRI in clinical practice are provided as a supplementary material (“Flowcharts” file).

9. Conclusion

Brachial plexopathy represents a diagnostic puzzle that typically needs multiple assessment tools, including electrodiagnostic and imaging studies. Ultrasound allows visualization of almost the entire postganglionic portion of the plexus and can show specific findings in different plexopathies that can impact diagnosis and management. Moreover, evidence of moderate-to-high certainty supports its use as a diagnostic adjunct, particularly when combined with electrodiagnostic studies. Continued standardization of scanning protocols and larger multicenter validation studies will further define its role in clinical practice. Future research directions include exploring the role of brachial plexus ultrasound in other disorders like amyotrophic lateral sclerosis (Yamazaki et al., 2025), and polyradiculopathy, and investigating the feasibility of elastography and deep machine learning (Yang et al., 2025).

Funding

This work did not receive any grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethical publication statement

We confirm that we have read the Journal's position on issues involved in ethical publication and confirm that this article is consistent with those guidelines.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material 1

Supplementary material 2

Supplementary material 3