This article presents a protocol for nerve ultrasound in polyneuropathies to aid the diagnosis of inflammatory neuropathies.
Nerve ultrasound is increasingly used in the differential diagnosis of polyneuropathy as a complementary tool to nerve conduction studies. Morphological alterations of the peripheral nerves, such as increasing the cross-sectional area (CSA), have been described in various immune-mediated polyneuropathies. The most prominent morphological changes in nerve ultrasound have been described for the chronic inflammatory demyelinating polyneuropathy (CIDP)-spectrum disease. CIDP may be distinguished from hereditary and other polyneuropathies by measuring the extent and pattern of nerve swellings (CSA increase). Typical findings in demyelinating inflammatory neuropathies are multifocal nerve swellings with inhomogeneous fascicular structure, while CSA increase in demyelinating hereditary neuropathies occurs in a more generalized and homogenous manner. In other non-inflammatory axonal neuropathies, nerves can appear with normal or slight CSA increases, especially in typical entrapment sites. This article presents technical requirements for nerve ultrasound, an examination procedure using a standardized examination protocol, current reference values for the CSA, and typical sonographic pathological findings in patients with inflammatory neuropathies.
Next to clinical examination, evaluating any large-fiber polyneuropathy includes an electrophysiological examination to characterize the motor or sensory system's involvement and differentiate axonal from demyelinating damage1. In axonal polyneuropathy, toxic and diabetic neuropathy are the leading causes, while in demyelinating polyneuropathies, hereditary or inflammatory neuropathies such as CIDP should be considered2,3,4. Commonly used diagnostic criteria for CIDP are the European Federation of Neurological Societies/Peripheral Nerve Society (EFNS/PNS) criteria established in 2005 and revised in 2010 and 20215. These define clinical and electrophysiological criteria to diagnose CIDP and describe additional criteria such as nerve biopsy to detect demyelination or inflammation. However, in some cases, despite a thorough diagnostic workup, the cause of neuropathy remains ambiguous. In these cases, nerve ultrasound offers a complementary method to examine the nerves not functionally but morphologically6. Several studies proved the use of nerve ultrasound as an additional tool in diagnosing CIDP, so that the 2021 revised EFNS/PNS criteria implemented nerve ultrasound in the guideline5. The advantage of nerve ultrasound compared to other imaging methods such as magnet resonance neurography (MRN) is that it can be used directly by the treating neurologists as a bedside tool; it is relatively cost-efficient. It can be used repeatedly, as it is noninvasive and not painful.
Typical characteristics of CIDP observed in nerve ultrasound are cross-sectional-area (CSA) increase7,8, also found in hereditary polyneuropathies. In CIDP, this affects individual nerve segments heterogeneously7,9.
A variety of examination protocols have been published10,11,12,13,14,15 trying to clarify normal CSA values and determine the adequate anatomical positions of ultrasound examination. Some of these positions are similar in most examination protocols. However, a widely accepted protocol to standardize the examination process and simplify the interpretation of the measurements does not exist.
This article demonstrates the nerve ultrasound examination using a standardized protocol for polyneuropathies, presents various reference values for the CSA, and shows typical pathological findings in patients with inflammatory neuropathies.
Technical requirements for nerve ultrasound
The neuromuscular ultrasound is performed in B-mode (Brightness mode, two-dimensional image with gray levels) using the compound imaging of the corresponding sonographic device6,16. Compound imaging enables electronic control of the piezoelectric elements in the sonic probe (transducer) to illuminate the target structure from different angles17. The ultrasound waves are reflected in several directions due to the histological structure of the peripheral nerves. As a result of the sound coming from different angles, a more significant part of the otherwise lost reflections gets back to the sound probe (receiver) and can generate images. For neuromuscular ultrasound, a high-resolution ultrasound probe with 18 MHz linear array transducer, for deeper nerves, an additional 12 MHz linear array probe (e.g., to display tibial and fibular nerve in the popliteal fossa) is used6,16. Transducers with lower frequencies result in reduced spatial and lateral resolution so that the differentiation of the nerve boundaries from the surrounding structures is less precise. The optimal settings can be kept constant using a preset for neuromuscular imaging provided by the manufacturer. During the examination, the image depth and the focus position must be adjusted to the structure to be examined and constantly adapted to the position of the nerve. The B-image gain and the depth-dependent gain can be adjusted for image optimization with uniform brightness. Blood vessels are often close to neural structures and are often used as landmarks to make the measurements at the same position. To depict their anatomical interaction and distinguish between nerves and vessels, it is also necessary to display the flow velocity and direction using pulsed Doppler and color-coded duplex sonography16,18. The pulse repetition frequency must be adapted to the expected low flow velocities in the blood vessels of the extremities, or the power Doppler must be selected for color-coding16.
Nerves reflect the ultrasound waves differently from different angles of incidence so that the sonographic image varies in echogenicity (anisotropy)16,19. The best image is achieved from an orthograde angle since the ultrasonic waves are reflected most strongly by the nerves in this angle. For avoiding artificial anisotropy or nerve deformity, the probe must therefore be held in a neutral position during the examination without applying additional pressure perpendicular to the nerves (Figure 1). The cross-sectional area (CSA) is measured within the thin, hyperechoic epineurium (Figure 2) to avoid alterations of the epinerval tissue in the measurement19. More details on technical ultrasound can be found in References6,16,17,18,19,20,21.
All examinations for this work were performed in compliance with institutional guidelines of the Ruhr-University Bochum, Germany.
1. Experimental preparations
2. Ultrasound examination
Each ultrasound laboratory should establish its CSA reference values by collecting data from the healthy local population, as specific ultrasound machines and examiner or population-dependent variables can lead to slightly different results in each laboratory. However, to indicate which CSA values can be considered normal, data from two leading German nerve ultrasound groups and a recent meta-analysis of all published reference values so far13,14,15,22,23 are summarized in Table 1. Reference values for patients studied under this protocol in our department are those by Kerasnoudis et al.22 (Table 1).
Typical findings in demyelinating inflammatory neuropathies are multifocal nerve swellings with inhomogeneous fascicles, while nerve swellings in demyelinating hereditary neuropathies occur more generalized and homogenous12,24. The histologic correlate of increased CSA is assumed to be acute inflammation and repeated de- and remyelination; however, this remains to be investigated7. In other non-inflammatory axonal neuropathies, nerves can appear normal or slightly increased in size, especially in typical entrapment sites25,26,27,28.
To simplify the interpretation of the results, the adjusted Bochum Ultrasound Score is suggested as a scoring system, which helps to distinguish chronic inflammatory neuropathies such as CIDP from non-inflammatory neuropathies.
The adjusted Bochum Ultrasound Score is calculated from the number of sites with significantly enlarged CSA of six of the above-described measurement sites: median nerve at the forearm, the median nerve at the upper arm, ulnar nerve at the forearm, ulnar nerve at the upper arm, the radial nerve at the upper arm and sural nerve at the calf. Examination of only these six sites will take ~15 min. Each of these six sites is scored with 1 point if the nerve shows pathological CSA enlargement on one or both sides of the body. Thus, the minimum score is 0 points, and the maximum score is 6 points. With this scoring system, if ≥2 points are assigned, the diagnosis of CIDP is possible with a sensitivity of ~53% and a specificity of ~83%, even if additional axonal damage in nerve conduction studies results in difficult detection by electrophysiological criteria.
Different groups have proposed other scoring systems to differentiate between neuropathies10,11,18,29,30. None of these scores is widely used. The adjusted Bochum Ultrasound Score is based on earlier publications which describe the Bochum Ultrasound Score10 derived from four measurement sites to distinguish CIDP from Guillain-Barré Syndrome and the Nerve ultrasound protocol30 derived from nine measurement sites to differentiate CIDP from MMN, MADSAM and vasculitic or paraproteinemic neuropathy. These different scores should be used according to the exact question. The adjusted Bochum Ultrasound Score was developed to diagnose CIDP if nerve conduction studies show possible CIDP defined by electrophysiological EFNS/PNS criteria5.
However, even if the adjusted Bochum Ultrasound Score only uses six nerve sites for calculation, still all other described nerve sites and the whole course of each nerve should be examined to detect focal lesions31 or exclude homogenous enlargement. In the case of homogenous nerve enlargement, hereditary neuropathy should be considered24. Scoring systems for homogeneity and alterations of the fascicular structure were described before and may aid in evaluating homogeneity8,24,32.
For ultrasound images of a healthy person, see Figure 4; for example, images from a CIDP patient, see Figure 5.
Figure 1: Examination of the median nerve at the wrist. To avoid artificial anisotropy or nerve deformity, the probe must be held in a neutral position during the examination without applying additional pressure perpendicular to the nerves. Please click here to view a larger version of this figure.
Figure 2: Measurement of the cross-sectional area (CSA). The cross-sectional area is measured within the thin, hyperechoic epineurium. Please click here to view a larger version of this figure.
Figure 3: Overview of measuring sites for CSA. Blue stars – median nerve, green stars – ulnar nerve, red star – radial nerve, pink star – vagal nerve, yellow stars – cervical roots and brachial plexus, white stars – fibular nerve, purple stars – tibial nerve, brown star – sural nerve. Please click here to view a larger version of this figure.
Figure 4: Example images of a healthy person of the six nerve sites used in adjusted Bochum ultrasound score. A – median nerve at the forearm, B – median nerve at the upper arm, C – radial nerve at the upper arm, D – ulnar nerve at the forearm, E – ulnar nerve at the upper arm, F – sural nerve at the calf. Please click here to view a larger version of this figure.
Figure 5: Example images of a patient with CIDP of the six nerve sites used in adjusted Bochum ultrasound score. A – median nerve at the forearm, B – median nerve at the upper arm, C – radial nerve at the upper arm, D – ulnar nerve at the forearm, E – ulnar nerve at the upper arm, F – sural nerve at the calf. Please click here to view a larger version of this figure.
Kerasnoudis et al.22 | Grimm et al.23 | Meta-Analysis by | ||||
Fisse et al.13-15 | ||||||
Nerve | Site | mean CSA | standard deviation | mean CSA | mean CSA | 95% CI |
(mm2) | (mm2) | (mm2) | (mm2) | |||
Median nerve | Wrist | 8.43 | 2.07 | 10.6 | 8.3 | 7.9 – 8.7 |
Forearm | 6.6 | 1.6 | 7.2 | 6.4 | 5.9 – 6.9 | |
Elbow | – | – | 9.2 | – | – | |
Upper arm | 8.4 | 2.87 | 9.1 | 8.3 | 7.5 – 9.0 | |
Ulnar nerve | Guyon loge | 5.16 | 1.03 | – | 4.1 | 3.6 – 4.6 |
Forearm | 5.46 | 1.26 | 5.9 | 5.2 | 4.8 – 5.7 | |
Elbow | 5.33 | 1.4 | 8.7 | 5.9 | 5.4 – 6.5 | |
Upper arm | 6.53 | 1.82 | 7 | 6.6 | 5.1 – 6.1 | |
Radial nerve | Upper arm | 3.26 | 1.52 | – | 5.1 | 4.0 – 6.2 |
Vagal nerve | Carotid sheath | – | – | 2.2 | 2.2 | 1.5 – 2.9 |
C5 | – | – | 2.4* | 5.6 | 4.6 – 6.7 | |
C6 | – | – | 3.4* | 8.8 | 7.4 – 10.3 | |
C7 | – | – | – | 9.5 | 8.0 – 10.9 | |
Brachial plexus | Intrascalene space | 30.93 | 10.82 | – | – | – |
Supraclavicular space | 46.13 | 18.27 | – | – | – | |
Fibular nerve | Fibula Head | 7.1 | 2.3 | – | 8.4 | 6.8 – 9.9 |
Popliteal fossa | 8.60 | 1.77 | 8.4 | 7.9 | 6.6 – 9.2 | |
Tibial nerve | Popliteal fossa | 8.43 | 2.68 | 23.2 | 25.9 | 17.5 – 34.4 |
Malleolus | 6.36 | 1.45 | 10.2 | 10 | 7.7 – 12.4 | |
Sural nerve | Heads of gastrocnemius muscle | 1.82 | 0.64 | 2.2** | 2.4 | 1.7 – 3.1 |
* Grimm et al.23 measured diameter, not the CSA for C5 and C6 root. | ||||||
** Grimm et al.23 measured sural nerve at the lateral ankle. |
Table 1: Reference CSA values for patients. Proposed reference values are based on the publication of Kerasnoudis et al.22, Grimm et al.23 and a recent meta-analysis by Fisse et al.13,14,15. Reference values for patients studied in our department under the protocol presented here are those by Kerasnoudis et al.22.
Nerve ultrasound is a helpful additional diagnostic tool in polyneuropathies. It can give information on the possible causes of polyneuropathy depending on the extent and pattern of nerve enlargement. Moreover, CSA alterations in the longitudinal disease course of patients with CIDP were described to correlate to clinical disease course and treatment response33,34,35,36.
Critical steps within the protocol
For obtaining reproducible results, consistent methodology and standardization of the examinations are essential37,38. Each examiner must consider deviations resulting from different ultrasound devices and local differences of demographics. To ensure the high quality and reproducibility of ultrasound measurements, specific training of the examiner is also necessary13,14,15,21,39.
Modifications and troubleshooting of the technique
Typical findings in demyelinating inflammatory neuropathies are multifocal nerve swellings with inhomogeneous fascicles33,40. Therefore, the measurement of CSA of specific nerve sites is necessary, but the whole nerve must be scanned. Also, evaluation of fascicular structure and echogenicity can help in inconclusive cases, as not only CSA increase of the entire nerve but also intrafascicular swellings as well as hypoechoic and hyperechoic nerves are found in CIDP. Hypoechoic nerves are considered to result from acute edema, while hyperechoic nerve result from fibrous remodeling40,41.
Limitations of the technique
There are anatomical limitations for nerve ultrasound, i.e., examination of cervical nerve roots can be difficult to impossible in patients with obesity and short neck. Also, imaging of proximal nerve roots of the lower extremity nerves of the lumbosacral plexus is not possible due to the limited penetration depth of the ultrasonic rays. An alternative method, evaluating these nerves, is possible by MRN42, but ultrasound is the more common method due to its spatial and temporal flexibility and cost-effective use43.
Significance with respect to existing methods
Nerve ultrasound is recommended as an additional and complementary tool in diagnosing polyneuropathies to evaluate nerve morphology. Standard diagnostic workouts, including nerve conduction studies and other tools such as cerebrospinal fluid analysis, should still be performed.
Future applications of the technique
For experts in polyneuropathies, nerve ultrasound is of interest for diagnosis in clinical routine and as it can give insights into possible pathophysiologic aspects, i.e., represent inflammation. Therefore, nerve ultrasound is a promising method not only in clinical use but also in neuromuscular research. Also, with increasing progress in ultrasound technology, future ultrasound characteristics such as shear wave elastography or vascularization of peripheral nerves may add further aspects in assessing polyneuropathies.
The authors have nothing to disclose.
We acknowledge the support from Ruhr-University Bochum for our research on neuromuscular ultrasound.
Affiniti 70 | Philips GmbH | n/a | with preset for neuromuscular ultrasound |
L18-5 linear array transducer | Philips GmbH | n/a | |
Ultrasound gel | C + V Pharma Depot GmbH | n/a |