Keywords
ultrasound - morphologic changes - upper extremities
Introduction
Ultrasound (US) examination is a valuable diagnostic tool in hand surgery. The
development of high-frequency (15–25 MHz) and ultra-high-frequency (30–70 MHz) US
probes in combination with sophisticated image processing allows for the
visualization of the smallest nerves and nerve bundles [1]
[2].
Ultrasonography is used to visualize and assess impairment as well as traumatic
injuries to nerves [3]. The examination helps
to identify and document the cause and extent of nerve lesions as well as nerve
transections [4]
[5]. Compared to MRI, US scans are demonstrably more cost-effective, offer
higher soft-tissue contrast, and allow for real-time examinations of the gliding
ability of nerves with a dynamic component [6].
Limiting factors of this diagnostic tool are bones and nerves at a depth of more
than 5 cm depending on the chosen probe [7].
Due to its availability and low cost, US examinations remain a very important tool
when serial examinations are needed. Signs of nerve damage include a hypoechogenic
pattern, swelling or absence of nerve bundles, and intraneural or perineural
scarring. Trauma to peripheral nerve trunks can result in nerve fiber loss to
varying degrees where complex pathophysiological changes (morphological and
metabolic) subsequently occur at the site of injury [8]. US scans can be implemented immediately post-trauma. This differs
from ENMGs, where pathological changes may be detected 14 days at the earliest
following an injury.
According to Sunderland [9], histologic changes
can be divided into five categories. Sunderland I describes a neurapraxia, which
sonographically presents as unremarkable. Sunderland II and III describe an
axonotmesis, which can sonographically show an increase in size and swelling of the
fascicles. Sunderland IV (epineurium intact) and V (complete lesion) correspond to a
neurotmesis and US shows a patchy or diffuse increase in echogenicity [10]. In this study, we focused on Sunderland
stage V nerve lesions, defined as an interruption of the epineurium, perineurium,
and endoneurium, without spontaneous regeneration potential. By definition, a stage
IV lesion develops following suturing of the nerve, which may then have regenerative
potential.
In case of a loss of nerve continuity, microsurgical repair is mandatory [10]. Surgical nerve suturing is a delicate
operation and belongs to the field of microsurgery. The suture should allow for the
nerve endings to touch (the ends “should kiss”) without any pressure or tension in
the sutured area. The suture is performed using the single button suture technique
with a very tiny thread (e. g., Nylon 9–0) and should be done solely in the
epineurium [11]. Usually, a group of fascicles
is repaired with as few buttons as possible but as many as necessary. In nerve
surgery, US can be used to locate normal or pathological anatomy [12]. During microsurgical suturing, the
enveloping structure (epineurium) is readapted, thereby creating continuity that can
be seen sonographically.
Postoperative sutures are inconspicuous on MR imaging but, depending on the size and
type of suture, appear as intraneural hyperechoic spots with comet tail artifacts on
US [11]. US scans can provide important
information regarding the actual state of the neurorrhaphy (e. g., size, extent, and
location of postoperative scarring), indicate potential neuromas (neuroma in
continuity, terminal neuroma) [13] and may be
useful in advance of a potential second surgery [14]. In the case of complete nerve lesions stump neuromas develop in
continuity with the edges of the nerve (round hypoechoic masses, can be displaced or
retracted from the site of injury) [15]
[16]
[17]. In
partial nerve lesions, a neuroma may develop along the injured nerve. US scans can
estimate the percentage of involved fascicles [17]
[18]
[19]. Traumatic neuromas are a reactive
hyperplasia of neuronal and fibrous tissue of the nerve sheath, which develops at
the end of a proximal nerve stump [20]. A
neuroma in continuity is defined as an axonotmetic injury in which the axons are
disrupted but the connective tissue is continuous (crush or stretching) [21].
US findings in traumatic nerve injuries are well known [7]
[12]
[13], but there is no study comparing US findings
following traumatic nerve injury and repair with a clinical focus on sensory and
motoric nerve regeneration. In this study, we wanted to obtain data on the outcome
of nerve regeneration following traumatic nerve lesions and nerve repair based on US
examinations as well as clinical and electrophysiological outcomes. The focus was on
describing several objective parameters in US corresponding to nerve regeneration,
which aid in the decision making process regarding the selection of the further
treatment after nerve trauma or nerve surgery at any time during regeneration. The
approach of this study was a prospective evaluation of US findings as well as
clinical sensory and motor function after traumatic nerve injury and nerve repair.
We wanted to answer the questions: Are there any correlations between post-traumatic
morphologic nerve changes with US findings and clinical nerve function after
surgery? Is there a time frame in which pathologic sonographic changes occur?
Methods
Study population
The objective was to establish a descriptive, prospective cohort study. Eligible
patients who had undergone surgery were selected. Another requirement for
participation in this dual center study (1 university tertiary referral center
Level A, 1 regional trauma center Level B; no differences in technological
standardization) was a minimum follow-up of 6 months following the surgical
procedure. The inclusion criteria and examination protocol were standardized in
both centers. The study was approved by the respective local ethics committee.
Patient enrollment commenced on the 6/1/2017. Participants provided written
informed consent. The patient characteristics of the study participants are
shown in [Table 1]. The time baseline was
defined as the date of the traumatic nerve lesion. The inclusion criteria were a
traumatic nerve injury followed by a nerve repair (directly or with a nerve
graft) of the median, ulnar, or radial nerves at the level of the arm, treated
and examined during the period of 05/2017–05/2020 with at least 50% transection
of the truncal nerve. The exclusion criteria were an age below 18 years and
brachial plexus injuries. All participants were treated surgically and were
monitored by clinical and sonographic follow-up at 6 weeks, 3 months, 6 months,
9 months, and 12 months postoperatively. In addition, an ENMG examination was
arranged after 6 months and 12 months.
Table 1 Patient characteristics at
baseline.
|
Characteristic
|
Nerve lesions (n=20)
|
|
Sex distribution: Male sex (percentage of male sex)
|
18 (90%)
|
|
Age, y, median (IQR)
|
17–72 (42)
|
|
Baseline date
|
6/1/2017 to 12/29/2020
|
|
Left side
|
12 (60%)
|
|
Percentage of nerve lesion
|
17 x 100%; 2 x 95%, 1 x 50%
|
|
Type of included nerves
|
8x ulnar nerve, 12x median nerve (17x distal forearm, 3x
middle forearm)
|
|
Number (N) of microsurgical repairs, Numbers of (N)
microsurgical reconstruction with graft
|
N=20 N=0
|
Clinical examination
Clinical examinations were established at predefined intervals (6 weeks, 3
months, 6 months, 9 months, and 12 months postoperatively) and were performed
using a protocol with clear instructions for both physicians and patients. The
clinical examination consisted partially of an inspection (hyposensitivity,
hyper-/hypo-hidrosis, hyper-/hypotrichosis, muscle atrophy, skin abnormalities,
swelling, nail deformation, etc.). The clinical examination focused on
sensitivity, as a subjective parameter, and motor ability, as an objective
parameter. The sensitivity was recorded according to the Medical Research
Council (MRC) and was therefore graded as S0–S4 [22] (S0=absence of sensitivity in the autonomous area; S1=recovery of
deep cutaneous pain sensitivity within the autonomous area of the nerve;
S2=return of some degree of superficial cutaneous pain and tactile sensitivity
within the autonomous area of the nerve; S3/S3+= return of superficial cutaneous
pain and tactile sensitivity throughout the autonomous area, with disappearance
of any previous overresponse; S3+= return of sensitivity as in S3; but in
addition there is partial recovery of 2-point discrimination within the
autonomous area (7–15 mm); S4=complete recovery (2-point discrimination of
2–6 mm). This parameter was recorded every time following careful patient
instruction. The motoric ability was evaluated as an objective parameter using
the method of Janda, also according to the MRC, and therefore graded from M0-M5
(M0=no discernible muscle contraction: 0% muscle strength; M1=discernible
reaction, though not sufficient for movement: approx. 10% muscle strength;
M2=movement to the full extent, though not possible against gravity, i. e. in a
horizontal plane: approx. 25% muscle strength; M3=movement to the full extent
and against gravity, without additional external resistance: approx. 50% muscle
strength; M4=movement to the full extent against light to medium resistance:
approx. 75% muscle strength; M5=movement to the full extent, even against a
strong external resistance: 100% physiological muscle strength.). Pain was
assessed based on the Visual Analog Scale (VAS). In addition, the presence of a
Hoffmann-Tinel phenomenon was clinically assessed in each case.
Sonographic examination
The following parameters were visualized and evaluated during the sonographic
examination: Nerve cross-section, continuous fascicles, echogenicity, and
perineural scars. The nerve cross-section was measured in millimeters. For
comparison, the opposite nerve was measured as well. The number of continuous
fascicles was recorded sonographically. The echogenicity and peri- as well as
intraneural scars were assessed. Hypoechogenicity of the suture was defined as a
decrease in the echogenicity of the neural structure compared to the area
proximal to the suture. A perineural/intraneural scar was defined as an increase
in the echogenicity of the neural structure perineurally/intraneurally compared
to the area proximally.
Endpoints
The primary endpoint was to determine the morphologic sonographic changes of the
nerves including the response of the surrounding tissue after nerve repair. As a
secondary endpoint, any correlation between morphologic sonographic changes and
nerve function was assessed. The nerve cross-sectional area (CSA in
mm2), number of traversing fascicles, potential
hypoechogenicity/hyperechogenicity and presence of peri-/intraneural scars were
analyzed sonographically at 6 weeks as well as 3, 6, 9, and 12 months
postoperatively.
ENMG after 6 and 12 months
An ENMG examination was performed at 6 and 12 months. We used VikingTM
On Nicolet EDX (Nicolet-Viking-EDX.pdf (neuroswiss.ch)). The following
parameters were recorded: Motor and sensory nerve conduction velocity,
amplitude, and motor latency. An ENMG at 6 months was performed in all 20 cases.
After 12 months, an ENMG could be performed in all cases that still had a
follow-up (n=12).
Results
Participants
We included 20 nerve lesions in 18 patients with intraoperatively confirmed trunk
nerve injury. In the forearm, at least 50% of the cross-sectional area of all
nerve lesions was affected. 12 median and 8 ulnar nerves were involved. In 17
cases, nerve transection occurred in the distal forearm and in 3 cases in the
middle forearm. All lesions were treated surgically through direct nerve
suturing. 90% of the participants were men and the median age at the time of the
surgery was 42 years. There was a minimum follow-up of at least 6 months in all
cases (100%). A follow-up period of at least 9 months was achieved in 14 cases
(70%) and of at least 12 months in 12 cases (60%). In the clinical examination
at 12 months, sensitivity presented as follows: 1x S1, 5x S2, 5x S3, and 1x S4.
A motor function of M3–5 at 12 months was seen in 10 out of 12 patients. The
results of the clinical examinations are summarized in [Fig. 1] and [Fig. 2].
Fig. 1 Results of clinical examinations (sensitivity).
Fig. 2 Results of clinical examinations (motor function).
Ultrasound
The sonographic results are summarized in [Table
2] and [Fig. 3]
[4]
[5]
[6]
[7]
[8]. Compared to the average nerve cross-sectional area of 20
mm2 at 6 weeks postoperatively, the participants showed nerve
cross-sectional areas of 25.7 mm2, 26.3 mm2, 23
mm2, and 25 mm2 at 3, 6, 9, and 12 months,
respectively. Consequently, the average CSA in mm2 throughout the
follow-up period was more than 20 mm2. In comparison, the opposite
side showed average CSAs of 10.75 mm2 after 6 weeks. Compared to the
sonographically detected average number of 10 continuous fascicles at 6 weeks
postoperatively, participants had 10, 15, 12, and 11 continuous fascicles at 3,
6, 9, and 12 months, respectively. Compared to 6 patients with hypoechogenicity
at 6 weeks postoperatively, 6, 3, 1, and 2 patients showed hypoechogenicity at
3, 6, 9, and 12 months, respectively. Compared to 1 patient with sonographically
proven perineural scarring at 6 weeks postoperatively, 2, 2, 3, and 1 patient
had sonographically proven perineural scarring at 3, 6, 9, and 12 months,
respectively.
Fig. 3 Traversing fascicles.
Fig. 4 Hypo-echogenicity.
Fig. 5 Neuroma in continuity.
Fig. 6 Increased CSA.
Fig. 7 Perineural scar.
Fig. 8 Intraneural scar.
Table 2 Results of US.
|
Time points
|
|
Morphology US
|
6 weeks (n=20)
|
3 months (n=20)
|
6 months (n=20)
|
9 months (n=14)
|
12 months (n=12)
|
|
CSA in mm2
|
20
|
25.7
|
26.3
|
23
|
25
|
|
Traversing fascicles (Number of fascicles)
|
10
|
10
|
15
|
12
|
11
|
|
Hypo-echogenicity (Number of nerves affected)
|
6
|
6
|
3
|
1
|
2
|
|
Perineural scars (Number of nerves affected)
|
1
|
2
|
2
|
3
|
1
|
|
Intraneural scars (Number of nerves affected)
|
0
|
0
|
0
|
0
|
0
|
There was a statistically significant correlation between the number of
continuous fascicles on sonography at 6 weeks and the level of sensitivity at 6
weeks postoperatively (0.018). For sonography at 3 months and level of
sensitivity at 3 months postoperatively, the p-value was 0.032. For sonography
at 6 months and level of sensitivity at 12 months postoperatively, the p-value
was 0.031. The only statistically significant difference concerning the motor
function was a p-value of 0.003 for sonography at 6 weeks.
There were no intraneural scars (hyperechogenicity within the nerve).
Irregular bulging of hypoechogenic tissue at the neurorrhaphy site was
interpreted as a pathologic sign, reflecting inadequate fusion of nerve edges
and a sign of a postsurgical neuroma formation.
According to our definition (axonotmetic injury in which the axons are severed
but the connective tissue is continuous after surgical suturing [21]), all nerves developed a neuroma in
continuity. Of these 20 nerve lesions, 3 (15%) were clinically relevant (pain
VAS>3/10) at 6 months. Allodynia was not present in any patient.
Sensitivity of at least S3 increased from 25% at 6 months to 36% at 9 months and
to 50% at 12 months during the indicated period. Motor function of at least M3
increased from 55% at 6 months to 57% at 9 months and to 83% at 12 months.
ENMG
After 6 months (n=20), motor latency ranged from 2.52 ms to 10.7 ms (mean: 5.15
ms, 7 lesions without measurable motor latency). The amplitudes reached values
between 0.1 mV and 11.5 mV (average: 2.95 mV, 7 lesions without measurable
amplitude). Sensory nerve conduction velocity was detected in 4 cases and
averaged 43.8 m/s.
After 12 months (n=12), motor latency ranged from 3.2 ms to 8.75 ms (mean: 5.6
ms, 12 lesions with measurable motor latency). The amplitudes reached values
between 0.3 mV and 6.3 mV (average: 2.6 mV, 12 lesions with measurable
amplitude). Sensory nerve conduction velocity was detected in 4 cases and
averaged 45.5 m/s (of which one lesion did not show any sensory nerve conduction
velocity at 6 months).
Lesions with a low motor latency or no motor latency at 6 months at all showed
lower clinical motor scores during the course of the study (n=4, no motor
latency, M0–1). 2 cases did not show worse clinical symptoms despite having no
motor latency (n=2, no motor latency, M3–4). The remaining lesions more or less
showed the expected correlation between EMG and function (n=14, the better the
motor latency, the higher the M value). After 12 months, 2 lesions showed a
rather low clinical value (M2 and M3) despite good ENMG results (motor latency
6.3 ms and 8.75 ms). All but one lesion (n=13) showed an increase in amplitude
in ENMGs at 12 months (compared with EMG at 6 months). A prolongation of motor
latency and a slowing of nerve conduction velocity was shown in all lesions. If
fasciculations and signs of degeneration or regeneration were found in the ENMGs
performed, such signs were always noted.
The results of the ENMGs of S3/4 and M4/5 are summarized in [Table 3]. The amplitudes decrease from 6
months to 12 months, the motor latency increases, contrary to the expectation
for nerve regeneration. We have observed that the values that were newly
derivable after 12 months (n=7) are rather low and therefore reduce the
calculated average.
Table 3 ENMG results for S3/4 and M4/5.
|
Time points
|
|
Amplitudes (average)
|
Motor latency (average)
|
Nerve conduction velocity (average)
|
|
S3/4 (6 months) n=5
|
2.5 mV (0.1 mV – 4.9 mV) (n=4)
(1 x not derivable)
|
4.39 ms (2.52 ms – 4.8 ms) (n=4)
(1 x not
derivable)
|
48 m/s (n=1)
(4 x not derivable)
|
|
S3/4 (12 months) n=7
|
1.86 mV (0.4 mV – 4.9 mV) (n=7)
|
5.83 ms (4.9 ms – 8.75 ms) (n=7)
|
43 m/s (n=1)
(6 x not derivable)
|
|
M4/5 (6 months) n=7
|
3.39 mV (0.4 mV – 11.5 mV) (n=6)
(1 x not
derivable)
|
5.15 ms (3.5 ms – 10.7 ms) (n=6)
(1 x not
derivable)
|
42 m/s (n=1)
(6 x not derivable)
|
|
M4/5 (12 months) n=6
|
2.93 mV (0.3 mV–6.3 mV) (n=6)
|
5.36 ms (3.2 ms–8.3 ms) (n=6)
|
42.5 m/s (42 m/s; 43 m/s) (n=2)
(4 x not
derivable)
|
Correlation between US and EMG
No significant correlation was found between US and ENMG. In all lesions, the
amplitude (mV) slightly decreased from 6 months postoperatively to 12 months
postoperatively due to huge standard deviations. Nerve conduction velocity
remained virtually unchanged. Motor latency remained+- constant, although a
slight increase in motor latency was also observed in some cases. These
observations correlated with the clinical findings but not with US
morphology.
Discussion
For the primary endpoint, we chose the sonographic morphological changes of the
sutured nerves including the response of the surrounding tissue after nerve repair.
We used any correlation between sonographic morphological changes and nerve function
for the secondary endpoint.
Our results suggest that there are corresponding post-traumatic sonographic
morphological changes in nerve fibers after traumatic injury. Therefore, a
correlation between the clinical and electrophysiological sequence and the
postoperative time and function is expected. However, sonography only represents the
morphological changes and not the function of a nerve.
We emphasize that sonography is the only method to repeatedly visualize the structure
of a nerve after surgical repair without the high cost of MRI.
Preoperative
In science, histological parameters are most often used as predictors of
peripheral nerve damage and regeneration [8]
[23]. However, they are not
suitable as an additional examination in patients following a traumatic
transection of a nerve. Sonography is a noninvasive examination method and
should be taken into account when diagnosing both traumatically and
non-traumatically induced nerve impairment [24]. Sonography, if available, is recommended for traumatic nerve
lesions even during follow-up and in order to correlate with ENMG and/or
clinical findings [25].
Postoperative/CSA
Our findings suggest that following surgical nerve suture, the cross-sectional
area of a nerve remains enlarged compared to the opposite side, even months
after surgery. These are normal findings in the sutured area of the nerve and
are not pathological [11]. Thus, they do
not correlate with nerve function. In reconstructive surgery of peripheral nerve
lesions, sonography allows for a reliable postoperative evaluation of the
continuity of a nerve (e. g., sufficient neurorrhaphy). On a postoperative US
scan, the fascicles, which are well adapted by a suture, appear in continuity.
Perineural collections can be detected as a mild and fusiform increase in nerve
size (cross-sectional area).
Previous studies have also demonstrated that pathologic US findings do not
significantly correlate with ENMGs or the clinical picture [26]. Consistently large postoperative CSAs,
which persist for months, do not correlate with improved or impaired nerve
function. Almost all of the nerves examined in our study met the criteria for a
neuroma in continuity, but few were actually clinically relevant. The nerve
cross-sectional area shows a marked increase after a few weeks and persists for
up to 12 months.
The literature suggests that there is probably no direct correlation between
neuroma size, nerve function, and potential for recovery. By definition, all
lesions develop a neuroma during the course of recovery, but only 3 out of 20
cases were symptomatic in our cohort. However, if the neuroma is very large
(>5x the normal nerve diameter), regeneration is unlikely [10]. This extent of neuroma was not observed
in our study.
Previous data shows that an asymptomatic neuroma in patients with thyroid cancer
following neck dissection needed no therapeutic intervention. Even after two
years there were no significant changes in clinical or sonographic findings
[27].
Number of continuous fascicles
After the traumatic transection of a nerve, the distal segment undergoes a slow
process of degeneration (Wallerian degeneration) [8]. Our study shows that as early as 6 weeks after microsurgical
repair, the first “continuous” fascicles can be detected sonographically.
Surgery generates a neuroma in continuity here. We interpret this in the context
of Waller’s degeneration taking place. It can be assumed that the enveloping
structure of the nerve remains and sonographic differentiation is not
possible.
Hypo-echogenicity
Our study supports hypoechogenic changes starting early and then decreasing up to
12 months postoperatively. Irregular bulging of hypoechoic tissue at the
neurorrhaphy site should be interpreted as a pathological sign, representing
inadequate fusion of nerve edges and a sign of postsurgical neuroma formation
[28]. Reactive focal swelling related
to edema and venous congestion could also be the result of neural distortion by
the scar. Our results suggest that hypo-echogenicity occurs early in the
postoperative course and decreases markedly after 9 months.
Perineural/intraneural scarring
Our results showed that perineural scarring after neurorrhaphy was fortunately
minimal. The presence of fibrous tissue can hinder MR and US imaging. It can
encase the nerve, compressing the fascicles, and as such the nerve is no longer
visible in the scar tissue [11]. Therefore,
the question is thus how pronounced is the scarring postoperatively. Intraneural
scarring was never observed in our sonographic follow-up.
Neuroma
End-to-end neurorrhaphy is an established method that significantly reduces the
incidence of symptomatic neuromas [29].
Misdirected or off-target outgrowth, excessive scar tissue, or a gap between the
nerve stumps prevents the axons from reaching their final destination. Such an
unsuccessful regeneration process leads to the formation of a neuroma [21]. Neuromas in continuity can only be
evaluated in conjunction with clinical, sonographic, and ENMG findings.
Correlation of US, ENMG, and clinical findings
No significant correlation was found between US, ENMG, and clinical findings. In
all lesions, the amplitude (mV) increased from 6 months postoperatively to 12
months postoperatively. Nerve conduction velocity remained slightly decreased.
Motor latency improved for the most part, although an increase in motor latency
was also observed in some cases. No statistically significant correlation
between ultrasound and clinical findings was observed. However, lesions with a
low motor latency or no motor latency at all at 6 months did show a lower
clinical motor score during the course of the study. Still, some cases did not
show worse clinical symptoms despite worse motor latency. The predominant
lesions more or less showed the expected correlation between ENMG and function
(more motor latency, higher M value). It shows that although in many cases an
ENMG examination provides a supporting and correlating tool, there are also
cases where this correlation is absent. We interpret this in our study based on
the distribution of the data and the increase in inducible stimuli (increase
from n=4 to n=7). We have observed that the values that were newly derivable
after 12 months are rather low and therefore reduce the calculated average. It
can therefore be assumed that during later regeneration these values will worsen
the expected average value and thus influence our results.
US findings in traumatic nerve injuries are well known [7]
[12]
[13], but in contrast to
other already published studies, our study compares US findings following
traumatic nerve injury and repair with the clinical outcome regarding sensory
and motoric nerve regeneration. In the literature, US is recommended for the
diagnostic workup of peripheral nerve lesions in addition to clinical and
electrophysiological investigations and should be used in the clinical workup of
traumatic nerve lesions. With this study, we provide initial indications of the
corresponding temporal sequences of morphological changes and possible
correlations between the individual clinical findings.
Limitations
The literature shows that depression and the workers' compensation status
are significantly associated with the formation of symptomatic neuromas [29]. Such factors were not considered in
this study. Furthermore, only 20 lesions could be included, which limits the
power of the study.
Conclusion
To our knowledge, this is the first study comparing US findings after traumatic nerve
injury and nerve repair with clinically examined sensory and motoric nerve
regeneration and corresponding ENMG findings. It supports the presence of
post-traumatic morphological changes in nerve fibers seen with US scans. These
post-traumatic morphological changes may be observed in a correlating sequence with
postoperative function over time. In accordance with this, our data suggests that
the more continuous fascicles are present in the US examination at 6 months
postoperatively in mixed motor and sensory nerves, the more likely sensitivity will
return at 12 months. Finally, in this prospective study, we only evaluated 20 nerve
lesions, highlighting the interest in future studies that quantify the presence of
post-traumatic sonographically detectable morphological changes in nerve fibers
after traumatic injury to a greater extent. Our data provides pioneering data for
future required studies.
Bibliographical Record
Léna G. Dietrich, Bettina Juon, Christian Wirtz, Esther Vögelin. Ultrasonographic Evaluation of Morphological Changes in Peripheral
Nerves after Traumatic Injury and Nerve Repair – A Prospective
Study. Ultrasound Int Open 2024; 10: a23786902.
DOI: 10.1055/a-2378-6902
