|Year : 2021 | Volume
| Issue : 3 | Page : 151-157
The role of electrophysiology in the diagnosis of radiculopathy and its comparison with magnetic resonance imaging
Ebru Boluk1, Cansu Uzunoğlu2, Yaprak Seçil1, Yusuf Kenan Çetinoğlu3, Şehnaz Arıcı1, Mustafa Fazıl Gelal4, Tülay Kurt-İncesu1
1 Department of Clinical Neurophysiology, Katip Çelebi University Atatürk Training and Research Hospital, İzmir, Turkey
2 Department of Neurology, Katip Çelebi University Atatürk Training and Research Hospital, İzmir, Turkey
3 Department of Radiology, Batman Training and Research Hospital, Batman, Turkey
4 Department of Radiology, Katip Çelebi University Atatürk Training and Research Hospital, İzmir, Turkey
|Date of Submission||18-Jan-2021|
|Date of Decision||17-Apr-2021|
|Date of Acceptance||15-Jun-2021|
|Date of Web Publication||20-Sep-2021|
Department of Neurology-Clinical Neurophysiology, Izmir Katip Celebi University, Atatürk Training and Research Hospital Karabaglar, Izmir
Source of Support: None, Conflict of Interest: None
Objectives: Patients are frequently referred to electrophysiology laboratories by physicians from various specialties with a presumptive diagnosis of radiculopathy. Recent advances in spinal imaging techniques have led to an increase in magnetic resonance imaging (MRI) studies in the diagnosis of radiculopathy. This study aimed to compare these two techniques in the diagnosis of radiculopathy. Subjects and Methods: Data of 170 patients who were diagnosed as having radiculopathy using electrophysiologic studies in our laboratory by defining a myotome level and who had spinal MRI examinations in our hospital database were included in the study. Results: Among the 170 patients in our study, 40% (n = 68) had cervical radiculopathy and 60% (n = 102) had lumbar radiculopathy. In the electrophysiologic diagnosis of radiculopathy, the most sensitive root level of electromyography (EMG) was the C7 radix level in the cervical region and the L5 level in the lumbar region. Correlations between radiologic grading based on foraminal assessments for radicular involvement at the cervical region using MRI and EMG data showed fair agreement in C5, C7, and C8 radix levels (κ: 0.21–0.40), and the lumbar region showed fair agreement in L4 and L5 radix levels (κ: 0.21–0.40). Conclusion: Although EMG has indisputable importance in the diagnosis of radiculopathy, its use with MRI significantly increases the diagnostic power at most spinal levels.
Keywords: Cervical radiculopathy, electromyography, foraminal stenosis, lumbar radiculopathy, magnetic resonance imaging
|How to cite this article:|
Boluk E, Uzunoğlu C, Seçil Y, Çetinoğlu YK, Arıcı &, Gelal MF, Kurt-İncesu T. The role of electrophysiology in the diagnosis of radiculopathy and its comparison with magnetic resonance imaging. Neurol Sci Neurophysiol 2021;38:151-7
|How to cite this URL:|
Boluk E, Uzunoğlu C, Seçil Y, Çetinoğlu YK, Arıcı &, Gelal MF, Kurt-İncesu T. The role of electrophysiology in the diagnosis of radiculopathy and its comparison with magnetic resonance imaging. Neurol Sci Neurophysiol [serial online] 2021 [cited 2021 Oct 18];38:151-7. Available from: http://www.nsnjournal.org/text.asp?2021/38/3/151/326286
| Introduction|| |
Radiculopathy is a clinical condition that occurs when the spinal nerve root is affected for different reasons. The most common etiologic causes include disc herniations, spinal stenosis, degenerative changes, infections, and metabolic-infiltrative-demyelinating processes. Advances in spinal magnetic resonance imaging (MRI) have led to the preference of this method to evaluate patients presenting with a presumptive diagnosis of radiculopathy. Electrodiagnostic examination consisting of electroneurography and electromyography (EMG) (this abbreviation will be shortened to EMG for the remainder of the manuscript) is important in the diagnosis of radiculopathy because it gives information about the level of radiculopathy, as well as the pathophysiology of the resulting process. EMG keeps its strategic role because it demonstrates the severity and course of the event. Frequent presumptive diagnoses encountered in patients directed to electrophysiology laboratories are cervical or lumbar radiculopathies. After electrophysiologic diagnoses, generally, patients are not followed up, and no information can be obtained about the radiologic-electrophysiologic correlation. This study aimed to evaluate whether patients diagnosed as having cervical or lumbar radiculopathy using EMG are also diagnosed as having radiculopathy radiologically. We also aimed to define the diagnostic capabilities of EMG and determine the levels where it was the most or least sensitive.
| Subjects and Methods|| |
This study included patients who were admitted to the clinical neurophysiology laboratory of Izmir Katip Çelebi University Atatürk Training and Research hospital between January and December in the same year, completed assessments, and had a presumptive diagnosis of isolated radiculopathy (cervical or lumbar, at least one or more radix levels) or additional diagnoses to radiculopathy (such as entrapment neuropathy or polyneuropathy). All data files were screened by three researchers from the electrophysiology laboratory. Patients with a presumptive diagnosis of isolated radiculopathy and those with radiculopathy and other accompanying presumptive diagnoses were separately evaluated. After EMG evaluations, patients without a radiculopathy diagnosis were not included in statistical evaluation. Patients who were diagnosed as having radiculopathy after the EMG evaluation with a determination of myotome level were included in the statistical evaluation.
The clinical data and symptoms of the patients were not able to be included because this study was retrospectively designed to compare laboratory findings (EMG and MRI).
Between January 2018 and December 2018, 8121 patients were admitted to our laboratory for assessments. Presumptive diagnoses were isolated radicular involvement in 609 patients and radicular involvement with another additional diagnosis in 237 patients (total: 846, 10.4% of total). Electrophysiologically, radicular involvement findings were present in 36% (n = 222, 2.7% of total) in the isolated radicular involvement group. Electrophysiologically, radiculopathy was present in 30% (n = 73) of 237 patients who were consulted with additional prediagnoses. A total of 295 patients had electrophysiologic findings of radicular involvement (3.6% of the total admissions). The availability and quality of spinal MRI examinations in these patients were checked through the hospital information and management system and the computer-based patient operating system used in our hospital. Out of 295 patients with EMG findings of radiculopathy, 85 patients who did not have MRI scans, 28 patients with a history of spinal surgery, eight patients with nondiagnostic MRI scans, three patients with spinal cord lesions, and one patient with spondylodiscitis were excluded. The remaining 170 patients' MRIs taken either 6 months before or after the electrophysiologic investigation were accepted for the study.
This retrospective study was conducted by following the Declaration of Helsinki and approved by the local ethics committee (Date: 13/11/2019, No.: 467).
A standard protocol was performed on all patients who were admitted with a presumptive diagnosis of radiculopathy. The American Association of Electrodiagnostic Medicine guidelines recommend that for the optimal evaluation of a patient with suspected radiculopathy, needle EMG studies of a sufficient number of muscles, and at least one motor and one sensory nerve conduction study (NCS) in the relevant limb. In our laboratory, routine NCSs include median and ulnar sensory and motor NCSs for the upper extremity and peroneal and posterior tibial motor NCSs, and sural nerve sensory conduction studies for the lower extremity. Needle EMG for the assessment of cervical radiculopathy includes the biceps brachii (C5/C6), deltoid (C5/C6), triceps (C6/C7/C8), abductor pollicis brevis (C8/T1), and abductor digiti minimi (C8/T1) muscles. For lumbar radiculopathy, the rectus femoris (L2/L3/L4), tibialis anterior (L4/L5), peroneus longus (L5/S1), gastrocnemius medialis (S1/S2), extensor digitorum brevis (L4/L5/S1), and abductor hallucis brevis (L5/S1/S2) muscles are routinely evaluated. These standard methods were performed on all patients. In needle EMG examinations, at least two muscles innervated from the same root but with different peripheral nerves were studied to exclude a mononeuropathy pattern. Myotomes remaining at the proximal and distal areas of the affected myotome level were also examined to check for a wider root involvement. If the distinction between two levels was unclear in needle EMG, pronator teres (C6/C7), extensor indicis proprius (C7/C8), extensor digitorum communis (C7/C8), flexor digitorum sublimis (C7/C8), flexor carpi ulnaris (C8-T1), supraspinatus (C5/C6), rhomboid muscles (C4-C5), cervical paraspinal muscles for the upper extremities, tensor fascia lata (L4/L5/S1), peroneus longus (L5/S1), tibialis posterior (L5-S1), gluteus maximus (L5/S1/S2), and lumbar paraspinal muscles are also evaluated for the lower extremities. If more than one myotome was detected, each was recorded individually. In the electrodiagnosis of radiculopathy, the presence of denervation findings (fibrillation potentials and positive sharp waves) and/or chronic neurogenic motor unit action potential (MUAP) changes (long-duration, large-amplitude, polyphasic MUAPs) in at least two muscles innervated by different peripheral nerves from the same myotome or denervation potentials in paraspinal muscles were accepted with limb muscle findings related to the myotome. Radiculopathies in all phases (acute, subacute, and chronic) with myotomal distribution of neurogenic effects were included in the study. In addition to limb muscles, paraspinal muscles were also evaluated in selected patients for radiculopathy with entrapment neuropathy or radiculopathy with polyneuropathy. EMG devices used in our laboratory are the two-channel Medtronic Keypoint EMG (Skovlunde, Denmark) and two-channel Nihon Kohden EMG/EP measuring system (Tokyo, Japan).
Radiological technique and evaluation
Cervical and lumbar MRI scans were obtained using two different 1.5T MR scanners (GE Healthcare, Milwaukee WI, USA, and Siemens AG, Erlangen, Germany). The images were evaluated and findings were recorded by two experienced radiologists.
MR grading system for cervical neural foraminal stenosis
Cervical axial T2* Multiple Echo Recombined Gradient Echo (MERGE, GE) and Multi-Echo Data Image Combination (MEDIC, Siemens) images were evaluated. The acquisition parameters for MERGE and MEDIC sequences were TR: 674 ms, TE: 16 ms, slice thickness: 4 mm, NEX: 1, FOV: 135 mm × 180 mm, matrix: 512 × 512; TR: 510 ms, TE: 24 ms, slice thickness: 3.5 mm, NEX: 1, FOV: 135 mm × 180 mm, and matrix: 512 × 384, respectively. When evaluating cervical neural foraminal stenosis, the grading system proposed by Kim et al. was used. This system evaluates the association of the neural foramen with the cervical disc in axial T2-weighted images and assigns three grades according to the severity of stenosis:
Grade 0: No stenosis in the neural foramen.
Grade 1: The width of the narrowest part of the neural foramen is the same or narrower than the width of the extraforaminal radix; however, the neural foramen width is more than 50% of the radix width.
Grade 2: The width of the narrowest part of the neural foramen is the same or narrower than the width of the extraforaminal radix (<50%).
In addition to grading, the presence of contact with or compression to the spinal cord was evaluated and recorded.
A Grade of 1-2 and/or contact/compression of the spinal cord was considered as MRI positive [Table 1].
|Table 1: Statistical results of electromyography - magnetic resonance imaging correlation for cervical radiculopathy|
Click here to view
Magnetic resonance grading system for lumbar neural foraminal stenosis
Lumbar sagittal T1-weighted images were examined. Acquisition parameters for GE and Siemens scanners were TR: 674 ms, TE: 8 ms, slice thickness: 4 mm, NEX: 2, FOV: 280 mm × 280 mm, matrix: 512 × 512; TR: 794 ms, TE: 9 ms, slice thickness: 4 mm, NEX: 3, FOV: 280 mm × 280 mm, matrix: 320 × 320, respectively. The grading system proposed by Lee et al. was used to evaluate the lumbar neural foramen stenosis using nonfat suppressed T1-weighted images:
Grade 0: No stenosis in neural foramen.
Grade 1: Vertical or transverse obliteration of perineural fat without morphologic changes in the radix.
Grade 2: Both vertical and transverse obliteration of perineural fat without morphologic changes in the radix.
Grade 3: Morphologic change in the radix due to neural foramen narrowing.
In addition to the grading, the presence or absence of radix compression was evaluated at the level of the lateral recess and recorded.
Grade 2 or 3 and radix compression at the level of the lateral recess were considered to be significant [Table 2].
|Table 2: Statistical results of electromyography - magnetic resonance imaging correlation for lumbar radiculopathy|
Click here to view
Descriptive statistics of the data included mean, standard deviation, median, minimum, maximum, frequency, and rate. The Kappa significance test was used for the concordance analysis. Statistical analyses were performed usingthe Statistical Package for the Social Sciences version 25.0 (SPSS Inc., Chicago, IL, USA). For the interpretation of Kappa statistics, the adjustment levels suggested by Landis and Koch were used. Exact Clopper-Pearson confidence intervals were used to calculate confidence intervals for sensitivity and specificity.
| Results|| |
The data of 170 patients were re-evaluated and included in the study. Among them, 68 patients had cervical radiculopathy (two patients had bilateral cervical radiculopathy and two-sided MRI evaluation), and 102 patients had lumbar radiculopathy (18 patients had bilateral lumbar radiculopathy and two-sided evaluation).
The mean age in the cervical radiculopathy group was 51.94 ± 12.56 years. Of these patients, 51.5% (n = 35) were male, and 48.5% (n = 33) were female. The findings were located on the right side in 54.4%, at the left side in 42.6%, and bilateral in 2.9%. In the lumbar radiculopathy group, the mean age was 54.92 ± 14.3 years. Of these patients, 56.9% were male and 43.1% were female. The findings were on the right in 41.2%, the left in 41.2%, and bilateral in 17.6%.
In the cervical radiculopathy group, in addition to radiculopathy, five patients had polyneuropathy, eight patients had carpal tunnel syndrome, and four patients had ulnar neuropathy at the elbow.
Correlations between MRI evaluations based on both radiological grading of foraminal stenosis and the absence or presence of cord contact/compression at the cervical region and electrophysiologic findings were statistically significant for C7 and C8 roots but not for C5, C6, and T1 roots [Table 1]. Cord contact/compression with or without foraminal stenosis affecting any level was present in 49 of 68 (72.1%) patients who were electrophysiologically diagnosed as having cervical radiculopathy. There was no cord contact/compression at any level in 19 (27.9%) patients. The evaluation of patients with radiculopathy with cervical spinal MRI revealed that the most common radiologic findings were at C6 (73.3%) and C5 (46.3%) levels.
The evaluation of EMG findings of the patients detected as having radiculopathy in cervical MRI separately for each root level revealed that the most sensitive root level was C7. The levels at which radicular involvement was most commonly observed in the cervical area were C7 and C6 roots based on the electrophysiologic studies and C6 and C7 roots in radiologic studies, respectively.
The findings evaluated at the C5, C6, and T1 root levels were not statistically significant.
According to Cohen's kappa value at these root levels, a fair, slight, and poor level of agreement was found between EMG and MRI, respectively.
In a small number of patients with electrophysiologic C8 radiculopathy in EMG (n = 15), radiculopathy was not detected in 60% (n = 9) of the patients in the cervical MRI examinations.
The most frequent radiologic and electrophysiologic radiculopathy levels in the lumbar region were L5 and L4 spinal root levels. Radix compression at the lateral recess with or without foraminal stenosis affecting any level was present in 80 of 102 (78.4%) patients who were electrophysiologically diagnosed as having lumbar radiculopathy. Lateral recess compression was not detected in 22 patients at the lumbar level (21.6%). The evaluation of MRI scans of patients who were not detected as having radiculopathy in EMG revealed that the most common radiculopathy was at L5 (78.5%) and L4 (67.3%) roots radiologically [Table 2].
In our study, in radiologically diagnosed L5 radiculopathies, EMG revealed the correct level at a rate of 90.2%. This was followed by L4 and S1 radiculopathies, in which EMG showed the correct level at a rate of 66.7% and 53.7%, respectively. The findings evaluated at the L4 and L5 root levels were statistically significant and showed a fair agreement between EMG and MRI at these levels.
Although the number of radiologically diagnosed L2 radiculopathies was quite low, its EMG sensitivity was the lowest among all lumbar roots (sensitivity: 12.5%). EMG sensitivity for radiologically diagnosed L3 radiculopathy was 30.2%. The findings evaluated at the L2 and L3 root levels were statistically significant, and at these root levels, a slight agreement was found between EMG and MRI [Table 2].
| Discussion|| |
The radiology literature that we referenced for the radiologic grading of radiculopathy emphasized that the clinical equivalence of the classification was unknown and researchers suggested that it should be supported by new studies to elucidate the correlation between the radiologic grade and the severity of symptoms., In this study, we tried to understand the radiologic equivalent of electrophysiologic radiculopathy using the same grading system. We tried to evaluate whether patients who were diagnosed as having radiculopathy electrophysiologically also had this diagnosis radiologically.
We found that approximately 10% of patients admitted to an active EMG laboratory had a presumed diagnosis of radiculopathy, and 3.6% had a diagnosis of radiculopathy, electrophysiologically.
Cervical radiculopathies are most commonly seen at the age of 54 years and occur somewhat more frequently in men. The incidence of lumbosacral radiculopathies has been reported as 3%–5%. It mostly occurs between the third and fifth decades of life, and it affects both sexes equally. In our study, lumbar radiculopathy was more common in men, similar to that in cervical radiculopathy (56.9%). Although the mean age was higher in patients with lumbar radiculopathy, it was between 52–55 years in both groups. Right-side radiculopathy was detected at a slightly higher rate; the increased frequency at the right side can be associated with the right-hand dominance of the patients. The frequency of bilateral radiculopathy was significantly increased in the lumbar radiculopathy group. Accordingly, we can say that lumbar involvement is more likely to be bilateral than cervical radiculopathy.
Spinal MRI is successful in revealing the spinal cord, radixes, and their associations with intervertebral discs and vertebra, but it does not give information about the physiological integrity of nerves. EMG is complementary to imaging studies because it provides information about the functional status of the nerves.
NCSs alone are not sensitive to radiculopathies. Needle EMG is of great importance in the diagnosis of radiculopathy because of its ability to demonstrate both axonal loss and the compensatory reinnervation process in the involved myotome. F waves can be used in the diagnosis of radiculopathy but its value is limited.
The Quality Assurance Committee of the AAEM evaluated 50%–71% sensitivity of electrodiagnostic studies in the diagnosis of suspected cervical radiculopathy. It was found that there was a good correlation between needle EMG and radiologic data in 65% to 85% of patients; therefore, needle EMG examination showed moderate sensitivity and high specificity in the diagnosis of cervical radiculopathy.
Although spinal MRI provides clear information about the anatomic and structural details of the nerve roots, abnormal MRI findings are also frequently encountered in asymptomatic individuals. In our study, patients with normal electrophysiologic examinations in terms of radiculopathy at the cervical and lumbar levels but had varying rates of radicular involvement on MRI most frequently occurred in C6 (73.3%) and C5 (46.3%) roots in the cervical region and L5 (64.3%) and L4 (45.6%) roots in the lumbar region. In this condition, the most appropriate method would be to evaluate patients with clinical findings. We could not use the clinical data of patients because our study design was retrospective. This is one of the limitations of our study.
It should be kept in mind that EMG examinations may be normal if radiculopathies are in the acute period and/or have pure demyelinating characteristics. Various systems have been proposed in the literature for the radiologic grading of cervical foraminal stenosis.,,, We used the system proposed by Kim et al. We found the highest cervical radiculopathy rates at C7 and C6, similar to the literature., The most sensitive level of EMG was the C7 level. Considering that radiologic evaluation of patients who had no radiculopathy findings for C7 in EMG revealed the involvement of C7 in 39.1% and involvement of C6 level in 73.3%, this may be related to the false positivity of MRI at these levels. Another possible reason may be that less than the optimal number of muscles is involved in the electrophysiologic screens. Therefore, we think that, to increase diagnostic power electrophysiologically, more muscles at this level, such as the pronator teres (C6–7) and flexor carpi radialis (C6–7) should also be included.
The number of radiculopathies detected radiologically and electrophysiologically in C8-T1 roots was the lowest among all roots. The fact that the C8-T1 innervated distal muscles (APB, ADM, and FDI), which are frequently affected in upper limb entrapment neuropathies, are open to microtraumas due to their location, might be considered as possible causes of electrophysiologic errors at this level.
According to Dillingham et al., when one of the examined muscles was a paraspinal muscle, the diagnostic sensitivity of 94%–99% was detected for six muscles for cervical radiculopathy diagnosis. The rate decreased to 84%–92% for five muscles if the paraspinal muscle was not studied. One of the limitations of our study is that paraspinal muscles are not routinely studied in our daily practice.
Due to the retrospective nature of the study, the evaluation of axial cervical MRI images obtained from 4 mm thick sections used in daily practice instead of volumetric thin section images may be a reason for false negativities in MRI at cervical levels. The system recommended by Kim et al., which we used in our study, does not involve the T1 level. At this level, the different angle of the T1 root in the neural foramen from the upper levels, and the performance of our image analysis using axial sections may explain the low EMG-MRI concordance. Myotome overlaps between adjacent segments are not uncommon, and this complicates the localization of the single root. Sometimes, discrimination in EMG can be difficult due to overlaps in C5–C6 and C6–C7 levels in the cervical region. In a similar study evaluating the correlation between EMG and MRI findings, the correlation was significant in lumbar radiculopathy (P = 0.007) but not in cervical radiculopathy (P = 0.976).
Various systems have been described for grading lumbar foraminal stenosis.,, We used the system proposed by Lee et al. by modifying it to the S1-S2 level to evaluate the S1 root in MRI. The most sensitive level of EMG in the diagnosis of lumbar radiculopathy was the L5 root level. The most common false-positive results, in which EMG demonstrated findings of radix involvement, but MRI was normal in the lumbar region, were for L4 and S1 roots. The more oblique course of S1-S2 neural foramen and oblique course of the S1 nerve root may explain the discordance between the EMG and MRI findings related to S1 radiculopathy. This may also account for the patients who had negative MRI but positive EMG findings in the S1 root (n = 38). Distal foot muscles, which are prone to trauma, are routinely studied to screen radiculopathy in the S1 root, which might have resulted in false positivity by causing chronic neurogenic changes. In addition to these distal foot muscles, it may be useful to include the gluteus maximus muscle in routine study protocols to determine electrophysiologic S1 radicular involvement.
In our study, the number of cases with radiologically diagnosed L2 radiculopathy was low (n = 12), and electrophysiologic sensitivity was found to be the lowest at this level among all lumbar roots (sensitivity: 9.1%). It is often almost impossible to localize upper lumbar radiculopathies individually due to the overlap of rich neural networks in the L2, L3, and L4 myotomes. Rectus femoris, vastus lateralis, vastus medialis, and adductor longus muscles receive almost equal innervation from L2-4 myotomes. Therefore, it is difficult to reduce lumbar radiculopathies at these levels to a single root level.
There is a wide range of opinions about the optimal number of muscles for the electrophysiologic diagnosis of lumbosacral radiculopathies., The general recommendation is that the optimal number of muscles for electrodiagnostic evaluation in both cervical and lumbosacral radiculopathies can be considered as six, representing all root levels, one of which is a paraspinal muscle.
| Conclusion|| |
In this study, when EMG and MRI findings were compared without clinical data, a low level of concordance was found in L4-L5 radiculopathies for the lumbar region and C5, C7, and C8 radiculopathies for the cervical region. This is significant because these findings show that electrophysiologic or radiologic examinations alone are not sufficient to diagnose radiculopathy in the absence of clinical data and examination findings.
The causes of the electrodiagnostic error include the difficulty of electrophysiologically reducing radiculopathy to a single root level due to the intensity of myotomal overlaps, the absence of routine inclusion of paraspinal muscles during the electrophysiologic evaluation, and overdiagnosis of chronic neurogenic changes shown by the distal limb muscles in routine study protocols. Under such conditions, both methods have limitations. It is still valid that EMG and MRI complement each other together with the clinical picture.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Li JM, Tavee J. Electrodiagnosis of radiculopathy. Handb Clin Neurol 2019;161:305-16.
So YT, Weber CF, Campbell WW. American Association of Electrodiagnostic Medicine, American Academy of Physical Medicine and Rehabilitation Practice parameter for needle electromyographic evaluation of patients with suspected cervical radiculopathy: Summary statement. Muscle Nerve 1999;22:209-11.
Kim S, Lee JW, Chai JW, Yoo HJ, Kang Y, Seo J, et al
. A new MRI grading system for cervical foraminal stenosis based on axial T2-weighted images. Korean J Radiol 2015;16:1294-302.
Lee S, Lee JW, Yeom JS, Kim KJ, Kim HJ, Chung SK, et al
. A practical MRI grading system for lumbar foraminal stenosis. AJR Am J Roentgenol 2010;194:1095-8.
Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74.
Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 1934;26:404-13.
Radhakrishnan K, Litchy WJ, O'Fallon WM, Kurland LT. Epidemiology of cervical radiculopathy. A population-based study from Rochester, Minnesota, 1976 through 1990. Brain 1994;117 (Pt 2):325-35.
Tarulli AW, Raynor EM. Lumbosacral radiculopathy. Neurol Clin 2007;25:387-405.
Jensen MC, Brant-Zawadzki MN, Obuchowski N, Modic MT, Malkasian D, Ross JS. Magnetic resonance imaging of the lumbar spine in people without back pain. N Engl J Med 1994;331:69-73.
Park HJ, Kim SS, Lee SY, Park NH, Chung EC, Rho MH, et al
. A practical MRI grading system for cervical foraminal stenosis based on oblique sagittal images. Br J Radiol 2013;86:20120515.
Kang Y, Lee JW, Koh YH, Hur S, Kim SJ, Chai JW, et al
. New MRI grading system for the cervical canal stenosis. AJR Am J Roentgenol 2011;197:W134-40.
Park HJ, Kim SS, Han CH, Lee SY, Chung EC, Kim MS, et al
. The clinical correlation of a new practical MRI method for grading cervical neural foraminal stenosis based on oblique sagittal images. AJR Am J Roentgenol 2014;203:412-7.
Wildermuth S, Zanetti M, Duewell S, Schmid MR, Romanowski B, Benini A, et al
. Lumbar spine: Quantitative and qualitative assessment of positional (upright flexion and extension) MR imaging and myelography. Radiology 1998;207:391-8.
Levin KH. Approach to the patient with suspected radiculopathy. Neurol Clin 2012;30:581-604.
Yoss RE, Corbin KB, Maccarty CS, Love JG. Significance of symptoms and signs in localization of involved root in cervical disk protrusion. Neurology 1957;7:673-83.
Dillingham TR, Lauder TD, Andary M, Kumar S, Pezzin LE, Stephens RT, et al
. Identification of cervical radiculopathies: Optimizing the electromyographic screen. Am J Phys Med Rehabil 2001;80:84-91.
Levin KH. Electrodiagnostic approach to the patient with suspected radiculopathy. Neurol Clin 2002;20:397-421, vi.
Arslan Y, Yaşar E, Zorlu Y. Correlation of electromyography and magnetic resonance imaging findings in the diagnosis of suspected radiculopathy. Turk J Neurol 2016;22:55-9.
Attias N, Hayman A, Hipp JA, Noble P, Esses SI. Assessment of magnetic resonance imaging in the diagnosis of lumbar spine foraminal stenosis – A surgeon's perspective. J Spinal Disord Tech 2006;19:249-56.
Kunogi J, Hasue M. Diagnosis and operative treatment of intraforaminal and extraforaminal nerve root compression. Spine (Phila Pa 1976) 1991;16:1312-20.
Barr K. Electrodiagnosis of lumbar radiculopathy. Phys Med Rehabil Clin N Am 2013;24:79-91.
Tsao B. The electrodiagnosis of cervical and lumbosacral radiculopathy. Neurol Clin 2007;25:473-94.
Lauder TD, Dillingham TR, Huston CW, Chang AS, Belandres PV. Lumbosacral radiculopathy screen. Optimizing the number of muscles studies. Am J Phys Med Rehabil 1994;73:394-402.
Dillingham TR, Lauder TD, Andary M, Kumar S, Pezzin LE, Stephens RT, et al
. Identifying lumbosacral radiculopathies: An optimal electromyographic screen. Am J Phys Med Rehabil 2000;79:496-503.
Dillingham TR, Annaswamy TM, Plastaras CT. Evaluation of persons with suspected lumbosacral and cervical radiculopathy: Electrodiagnostic assessment and implications for treatment and outcomes (Part I). Muscle Nerve 2020;62:462-73.
[Table 1], [Table 2]