Clinical Neurophysiology 117 (2006) 392397 www.elsevier.

com/locate/clinph

Cervical nerve root stimulation. Part I: Technical aspects and normal data*
Steve Vucic, Kevin D. Cairns, Kristin R. Black, Peter Siao Tick Chong, Didier Cros *
Department of Neurology, Massachusetts General Hospital, 55 Fruit Street, Boston, GRB 1256 Boston, MA 02114, USA Accepted 12 October 2005 Available online 3 January 2006

Abstract Objective: Cervical nerve root stimulation (CRS) is a technique of assessing the proximal segments of motor axons destined to upper extremity muscles. Few studies report normal values. The objective was to determine CMAP onset-latencies and CMAP amplitude, area, and duration changes in healthy controls for the abductor pollicis brevis (APB), abductor digiti minimi (ADM), biceps, and riceps muscles. In addition, to determine the tolerability of CRS, as measured by the visual analog scale (VAS). Methods: We studied 21 healthy volunteers prospectively with CRS using four target muscles (APB, ADM, biceps, and triceps) bilaterally. Collision studies were used in all APB recordings. VAS was obtained in all subjects. Results: Mean CMAP onset-latencies were: APB 14G1.5 ms; ADM 14.2G1.5 ms; biceps 5.4G0.6 ms; triceps 5.4G1.0 ms. Onset-latency signicantly correlated with height for all nerves. The mean change in CMAP amplitude and area (%) between most distal stimulation and CRS was: APB reduction of 15.1G11.6 and 4.9G3.6%; ADM reduction of 21.1G10.7 and 17.2G8.8; biceps reduction of 10G11.5 and reduction of 8.7G6.8; triceps increase of 3.3G5.2 and 11.0G9.9% respectively. Mean CMAP duration change between most distal stimulation and CRS was: APB, increase of 20.4G7.4%; ADM, increase of 14.4G8.5%; biceps, increase of 13.9G10.8%; triceps, increase of 7.7G6.7%. The mean VAS score was 3.8G1.2, and all subjects completed the study. Conclusions: The present study establishes normative data and indicates that CRS is a well-tolerated technique. Signicance: The normal values may be used as reference data for the needle CRS technique in the assessment of proximal conduction abnormalities. q 2005 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: Cervical nerve root stimulation; Normal data

1. Introduction Cervical nerve root stimulation (CRS) is a motor conduction technique evaluating the entire length of peripheral motor axon by stimulating the neural elements in the intervertebral foramen and recording CMAP responses from target muscles (Barker et al., 1987; Berger et al., 1987). Several techniques have been used and include magnetic and electrical stimulation. Magnetic CRS utilizes a rapidly varying magnetic eld, produced by a magnetic coil placed over the lower neck area, to induce an electric
* Presented at the Biannual 10th Clinical Neurophysiology Workshop of the Australian Association of Neurologists, April 1923, Southport, Queensland, Australia. * Corresponding author. Tel.: C1 617 726 3642; fax: C1 617 726 2019. E-mail address: dcros@partners.org (D. Cros).

current (Cros et al., 1990; Mills and Murray, 1996). Electrical stimulation is achieved by either a monopolar needle inserted at the C5/C6 or C6/C7 vertebral level (Berger et al., 1987; Sander et al., 1999), or high voltage percutaneous electrical stimulation over the cervical vertebral column (Mills and Murray, 1986). Although magnetic CRS is virtually painless, supramaximal stimulation may not be achieved in one third of cases and the exact site of nerve excitation is uncertain (Cros et al., 1990). For the purpose of this paper CRS will refers to needle electrical stimulation of the cervical spinal roots. To date six studies have included normal values in studies using cervical root stimulation. Two of them (Lange et al., 1992; Mills and Murray, 1986) used the high voltage percutaneous method, while the other four (Berger et al., 1987; Cros et al., 1990; Menkes et al., 1998; Sander et al., 1999) employed the needle electrical method.

1388-2457/$30.00 q 2005 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2005.10.011

S. Vucic et al. / Clinical Neurophysiology 117 (2006) 392397

393

The aim of our study was to obtain normal data in the distribution of the median (APB), ulnar (ADM), radial (triceps), and musculocutaneous (biceps) nerves bilaterally, using the CRS technique for the following variables: (i) CMAP onset latency and side-to-side difference; (ii) CMAP amplitude and area with change between the distal to CRS sites of stimulation, and side-to-side difference; (iii) CMAP duration and change between distal CRS sites of stimulation, and side-to-side difference. We also reevaluated the site of activation of the motor bers by comparison of CRS and F-wave minimum latencies.

2. Methods 2.1. Subjects Twenty-one healthy volunteers (15 males and six females, mean age 36.8G15.2 years, 2366), were studied prospectively. The mean height of volunteers was 170G 10 cm. Written informed consent was obtained from each participant prior to any study procedures in accordance with Institutional Review Board (IRB) of the Massachusetts General Hospital. 2.2. Neurophysiologic technique Nerve conduction studies (NCS) were performed on an Oxford Synergy (Oxford Instruments, Old Woking, UK) electromyography machine. The temperature of the upper extremities was maintained at 32 8C. The median, ulnar, musculocutaneous, and radial motor nerves were studied bilaterally. Stimulation sites for the median nerve were at the wrist, elbow, and nerve root, recording from the abductor pollicis brevis (APB); ulnar nerve at the wrist, below elbow, above elbow, and nerve root, recording from the abductor digit minimi (ADM); musculocutaneous nerve at Erbs point (EP) and nerve root, recording from the biceps brachii; and radial nerve at EP and nerve root, recording from the triceps. During CRS, the subject was sitting comfortably in a chair with the neck in slight exion. The cathode used was a monopolar needle (20 gauge, 50 mm length, Oxford Instruments, Old Woking, UK), while the anode and the ground were 32 mm stainless steel disk electrodes (Oxford Instruments, Old Woking, UK). The cathode was inserted perpendicularly to the skin, at the C6C7 interspace, 1 cm lateral to the spinous process and advanced though the paraspinal muscles to the vertebral lamina in a parasagittal plane. The cathode was ipsilateral to the recording side, while the anode position was varied on the midline, 34 cm rostral, lateral, and caudal to the cathode, to ensure maximum CMAPs. At each anodal position, 34 CMAPs were obtained and the largest potential was recorded. Between 9 and 12 stimuli were delivered with CRS to ensure a maximum CMAP response. After each stimulus,

the depth of the cathode was readjusted for optimal results. For EP stimulation, a standard 10 mm surface electrode was used as the cathode positioned medially in the supraclavicular area, and a stainless steel 32 mm disk surface electrode (Oxford Instruments, Old Woking, UK) as the anode positioned over the scapula (Roth and Magistris, 1987). CMAPs were recorded using stainless steel 10 mm disk electrodes (Oxford Instruments, Old Woking, UK) in a belly tendon arrangement. Collision study was performed with CRS when recording the CMAP response over the APB muscle to eliminate ulnar nerve contribution (Kimura, 1976). Briey, collision studies were performed by applying a supramaximal stimulus to the ulnar nerve at the wrist, 7 ms before CRS. The anti-dromic ulnar nerve volley collided and cancelled out the ulnar nerve component generated with CRS. Stimulus intensity was set to 100 mA (300 V), pulse width 1 ms, so as to ensure supramaximal stimulation. The band pass was 210 kHz, with a sensitivity of 5 mV/division for determination of CMAP amplitude, and 500 mV/division for onset latency. Sweep speed was 5 ms/division. The ground electrode was positioned between the cathode and the recording electrodes over the ipsilateral shoulder. 2.3. Nerve conduction parameter analysis For each motor nerve and at each site of stimulation, the following parameters were recorded: CMAP onset latency (ms); baseline to negative peak CMAP amplitude (mV); CMAP duration and area of the negative phase (ms), and temporal dispersion [TD], e.g. total duration of the CMAP measured from the initial negative deection of the CMAP to the last negative baseline crossing (e.g. the end of the last negative peak). In addition, a side-to-side difference was established for each parameter. The percent drop in CMAP amplitude between proximal (nerve root) and distal (wrist) sites of stimulation, was calculated using the following formula: CMAP (distal)-CMAP (proximal)/CMAP (distal) multiplied by 100. 2.4. Stimulation site relative to F response turn-around time. Data obtained from the ADM and APB muscles were studied to localize the site of excitation of the fastest conducting motor bers with CRS. We localized this site relative to the F response turn-around point. For this, we calculated the peripheral motor conduction time (PCT) to ADM and APB muscles using the formula: PCT Z Flat C DML K1=2 (Cros et al., 1990). In this formula, Flat is the minimum F response latency of 10 consecutive F responses, DML is the distal motor latency, and one is the 1 ms estimated turnaround time at the anterior horn cell. The site of excitation of the motor axons was derived by the following formula: DZ[PCTKCMAP onsetCRSlatency] multiplied by CV. Where D is the distance away from the cell

394

S. Vucic et al. / Clinical Neurophysiology 117 (2006) 392397

body and CV is conduction velocity, which is assumed to be 60 m/s in the proximal segments of the peripheral motor bers (Mills and Murray, 1986). 2.5. Discomfort and side-effects Subject discomfort was assessed with a 10 point visual analog pain scale (VAS), 0 being no pain and 10 being the worst possible pain at the completion of CRS. 2.6. Statistics The results obtained on both sides for each target muscle in a given subject were averaged and expressed as meanG standard deviation (SD). Side to-side differences for different neurophysiologic parameters in each nerve were compared using the paired t-test. Data followed a normal distribution as indicated by the fact that the mode, median, and mean were similar for all variables tested. The relationships between height and onset latency for all nerves were assessed with Pearsons correlation coefcient. AP!0.05 was deemed signicant.

The CMAP onset latency signicantly correlated with height for target muscle: APB, rZ0.75; ADM, rZ0.75; biceps, rZ0.78; triceps, rZ0.77 (P!0.05). 3.2. CMAP amplitude and area APB muscle. The mean CMAP amplitude with wrist stimulation was 12.5G3.0 mV (7.718.9), while with CRS it was 10.6G2.7 mV (6.316.8), the mean reduction being 15.1G11.6% (039), and mean side-to-side difference of 8.7G7.5% (026, PZ0.21). The mean reduction in CMAP amplitude between elbow and CRS sites of stimulation was 11.9G10.6% (039), with mean side-to-side difference of 12.8G12.8% (039, PZ0.33). The mean CMAP area with wrist stimulation was 39.5G 7.5 mV ms (27.558.7) while with CRS 37.9G9.3 mV ms (23.559.3). Mean reduction in CMAP area was 4.9G3.6% (013.8), with mean side-to-side difference of 15.9G8.5% (031). The mean CMAP area with elbow stimulation was 37.8 G7.5 mV ms (2856.3), while with CRS 38.5 G 9.4 mV ms (25.559.3), representing a 1.8G2.8% increase in CMAP area. The CMAP amplitude was reduced by collision studies in 15/21 (86%) subjects and 31/42 (73%) nerves, while in 6/21 (29%) subjects and 10/42 (24%) nerves the CMAP amplitude increased after collision (Table 2). 3.3. ADM muscle The mean CMAP amplitude with wrist stimulation was 11.6G2.5 mV (7.518.0) while with CRS it was 9.1G 2.3 mV (6.017.6). Mean reduction in amplitude was 21.1G10.7% (037) with mean side-to-side difference of 8.0G5.3% (019, PZ0.48). The mean reduction in CMAP amplitude between above-elbow and CRS sites of stimulation was 13.8G10.2% (036). Mean CMAP area with wrist stimulation was 36.0G 6.7 mV ms (23.749.1) while with CRS was 33.0 G 6.6 mV ms (1744.1). Mean reduction in area was 17.2 G8.8% (035) with mean side-to-side difference 10.0G5.9% (019, PZ0.35). The mean reduction in CMAP area between above-elbow and CRS sites of stimulation was 11.8G8.3% (030).

3. Results All neurophysiologic ndings are summarized in Table 1. Examples of waveforms are presented in Fig. 1(A)(D). 3.1. CMAP onset latency CRS to APB muscle. The mean CMAP onset latency was 14G1.5 ms (11.917.8), with mean side-to-side difference of 0.5G0.27 ms (0.11.1, PZ0.34). CRS to ADM muscle. The mean CMAP onset latency was 14.2G1.5 ms (11.617.6), with mean side-to-side difference of 0.4G0.4 ms (01.4, PZ0.39). CRS to biceps brachii. The mean CMAP onset latency was 5.4G0.6 ms (3.96.3), with mean side-to-side difference of 0.3G0.2 ms (00.8, PZ0.47). CRS to triceps. The mean CMAP onset latency was 5.4G 1.0 ms (3.77.3), with mean side-to-side difference of 0.7G 0.6 ms (01.7, PZ0.49).
Table 1 Summary of neurophysiologic variables Nerve-muscle Median-APB W CMAP onset latency (ms) CMAP Amplitude (mV) Area(mV ms) CMAP duration (ms) 12.5 39.5 5.6 CRS 14{1.5} 10.6(15.1) 37.9(4.9) 7.2(20.4) Ulnar-ADM W 11.6 36.0 6.0

Mscl-BB CRS 14.2{1.5} 9.1(21.1) 33.0(17.2) 7.2(14.4) EP 8.0 44 10.9 CRS 5.4{0.6} 7.9(10.0) 41(8.7) 12.4(13.8)

Radialtriceps EP 13.6 70.0 9.4 CRS 5.4{1.0} 14.1(3.3) 78.0(11) 10.1(7.7)

ADM, abductor digit minimi; APB, abductor pollicis brevis; BB, biceps brachii; CMAP, compound muscle action potential; CRS, cervical nerve root stimulation; EP, Erbs point; Mscl., musculocutaneous nerve; W, wrist stimulation; { }, standard deviation; ( ), mean percent difference between proximal and distal sites of stimulation.

S. Vucic et al. / Clinical Neurophysiology 117 (2006) 392397

395

A Wrist

Wrist

Below elbow Elbow Cervical root Above elbow Cervical root

5 mV 2 ms C Erbs Point D Erbs Point

Cervical root

Cervical root

Fig. 1. Compound muscle action potentials (CMAP) recorded over the (A) Abductor pollicis brevis, (B) Abductor digit minimi, (C) biceps brachii, (D) triceps muscles. The sites of stimulation are indicated next to each waveform. Although the mean data show that the CMAP amplitudes are higher with Erbs Point (EP) stimulation when recording over the biceps brachii, (C) and (D) demonstrate that the CMAP amplitude with cervical nerve root stimulation, was larger than with EP stimulation. This probably resulted from sub-maximal stimulation at EP despite modication in the stimulating technique. Clinically, submaximal stimulation at EP this may be important as proximal CB (between EP and nerve roots) may be missed.

3.4. Biceps brachii The mean CMAP amplitude with EP stimulation was 8.0G1.8 mV (5.012.2) while with CRS it was 7.9G 2.2 mV (4.113.2). Mean amplitude reduction was 10G 11.5% (038) with side-to-side difference of 11.3G10.8% (039, PZ 0.25). Mean CMAP area with EP point stimulation was 44.0G13.0 mV ms (20.070.0) and with CRS was 41.0 G10.6 mV ms (20.072.2). Mean area reduction being 8.7G6.8% (033) and side-to-side difference 6.4G5.8% (024, PZ0.48). In 15/21 (71%) subjects and 30/42 (85%) of nerves, the CMAP amplitude and area were greater with EP point stimulation than with CRS, while in 6/21 (29%) subjects and 7/42 (17%) nerves the CMAP amplitude and area were greater with CRS stimulation. 3.5. Triceps The mean CMAP amplitude with EP stimulation was 13.6G2.9 mV (9.119.9) and with CRS was 14.1G3.3 mV (1.619.9). Mean amplitude increased by 3.3G5.2% (020) with side-to-side difference of 4.7G4.9% (016, PZ0.34). The mean CMAP area with EP stimulation was 70.0G 16.0 mV ms (40.5109.3) and with CRS was 78.0 G 16.0 mV ms (46.7107.8). The mean area increase was

11.0G9.9% (030) and side-to-side difference of 11.2G 7.7% (030, PZ0.48). In 13/21 (62%) subjects and 26/42 (62%) nerves the CMAP amplitude and area were greater with EP stimulation than with CRS, while in 8/21 (38%) subjects and 10/42 (24%) nerves the CMAP amplitude and area were greater with CRS.

4. CMAP duration and temporal dispersion All CMAPs were synchronous with biphasic morphology (Fig. 1). None showed abnormal temporal dispersion.
Table 2 CMAP amplitudes recorded over the APB pre and post collision CMAP amplitude before collision (mV) Group 1 (nZ15) Group 2 (nZ6) 11.7G2.7 9.7G2.4 CMAP amplitude after collision (mV) 10.3G2.6 11.1G3.5 Change (%) K11.3G7.8 C11.3G11.8

CMAP, compound muscle action potential; group 1, refers to subjects in whom the CMAP amplitude recorded over the abductor pollicis muscle reduced after collision; group 2, refers to subjects in whom the amplitude increased post collision; n, number of patients.

396

S. Vucic et al. / Clinical Neurophysiology 117 (2006) 392397 Table 3 Values 3 SD from the mean Nerve CMAP latency (ms) 18.5 18.7 7.2 8.4 CMAP amplitude reduction (%) 49.9 53.2 44.5 18.9a CMAP area reduction (%) 15.7 43.6 29.1 40.7a CMAP duration increase (%) 42.6 39.9 46.3 27.8

4.1. APB muscle The mean CMAP duration with wrist stimulation was 5.6G0.6 ms (4.37.4) while with CRS it was 7.2G1.2 ms (5.011.3), with mean increase in the CMAP duration being 20.4G7.4% (030). 4.2. ADM muscle The mean CMAP duration with wrist stimulation was 6.0G0.8 ms (4.98.8) while with CRS 7.2G1.0 ms (6.0 9.3), with the mean increase in CMAP duration being 14.4G8.5% (032). 4.3. Biceps The mean CMAP duration with EP stimulation was 10.9G2.9 ms (5.818.8) while with CRS it was 12.4G 3.7 ms (5.320), with the mean decrease being 13.9G10.8% (038). 4.4. Triceps The mean CMAP duration with EP stimulation was 9.4G1.2 ms (6.412) while with CRS 10.1G1.3 ms (7.5 13.4), with the mean increase in CMAP duration being 7.7G6.7% (024.3).

Median Ulnar Musculocutaneous Radial

CMAP, compound muscle action potential; reduction, refers to reduction in the neurophysiologic parameter between distal (wrist or Erbs point) and proximal (nerve root) sites of stimulation. a Mean amplitude and area increased with proximal stimulation.

7. Discussion We report normal data obtained with CRS in a large series of controls, in whom we performed bilateral studies recording form APB, ADM, biceps and triceps. Normal values for latencies, amplitude, area and temporal dispersion (meanG3 SD) are shown in Table 3. We found that CRS is a safe and well tolerated nerve conduction technique. Although normal data was reported in other studies, there were major differences when compared to our study (Berger et al., 1987; Cros et al., 1990; Menkes et al., 1998; Sander et al., 1999). Berger et al. (1987) studied the APB, ADM, biceps and triceps bilaterally in 12 subjects, and reported maximum interside CMAP amplitude and latency asymmetries of 17% and 0.5 ms respectively (Berger et al., 1987). However, no information was provided on the range of CMAP onset latency, amplitude, area, and duration changes. Cros et al. (1990), reported a range of CMAP onset latency and amplitude change with side-to-side differences for the ADM, biceps and triceps in 10 subjects, which were similar to our ndings. However, the APB was not studied and CMAP area and duration changes were not reported. Menkes et al. (1998) evaluated the APB, ADM, biceps and triceps in 15 subjects and found that the maximum amplitude reduction was less than 50%. Most recently, Sander et al. (1999), reported that the maximum CMAP amplitude reduction in 15 subjects, for the APB and ADM was less than 50%. However, the range of CMAP area, duration, and latency change was not reported in either study (Menkes et al., 1998; Sander et al., 1999). Moreover, the CRS technique was different in that the monopolar needle (cathode) was inserted at the level of the C5 or C6 vertebral body in both studies. Collision studies should be used routinely when assessing the APB with CRS. This is supported by our ndings that in 18/21 (86%) subjects and 26/42 (62%) nerves the CMAP amplitude recorded over the APB muscle was reduced post collision due to elimination of the summated ulnar nerve component supplying the deep muscles of the thenar eminence. Thus, collision studies may unmask conduction block in the pathological setting.

5. Site of stimulation The average conduction time of the responses evoked to the ADM muscle by CRS was 0.12 ms longer than the PCT derived from the minimum F response latency, while the average conduction time to the APB was 0.24 ms longer. Using the mean motor conduction velocity of 60 m/s, the mean site of excitation is calculated to be 7.2 mm away from the cell body for the ulnar nerve and 14.4 mm for the median nerve anterior horn cell.

6. Discomfort and side-effects Each stimulus caused a sudden contraction of the cervical extensor muscles and shoulder abductor muscles, a feature that must be explained to all subject prior to performing the study. None of the subjects complained of severe discomfort, and the study was completed in all cases. The mean visual analog pain scale (VAP) was 3.8G1.2 (2.56). The most uncomfortable aspect of the procedure was the electrical stimulation with the monopolar needle. Two subjects experienced excessive sweating during the procedure and one subject had a pre-syncopal episode. There were no other side effects.

S. Vucic et al. / Clinical Neurophysiology 117 (2006) 392397

397

Conversely, in 6/21 (29%) subjects and 6/42 (14%) median nerves, the CMAP amplitude increased after collision studies. This may be explained by the elimination of the ulnar nerve contribution to the processes of desynchronisation and phase cancellation in the combined median and ulnar nerve pathways when recording over the APB muscle. The processes of phase interaction, summation and cancellation, may have inuenced CMAP variables in proximal muscles. Specically, the increase in the CMAP amplitude and area with CRS, when recording over the triceps, may be explained by either phase summation with CRS or sub maximal stimulation at EP. Collision studies, if possible, may have eliminated the effects of phase interaction. We previously reported that electrical stimulation at EP is frequently submaximal, (Cros et al., 1992; Cho et al., 2001). In this study, however, we did not nd any signicant differences in the mean CMAP amplitude between stimulation at EP and nerve root. In the majority of cases the CMAP amplitude was larger with EP stimulation. This may have resulted from the modication of the conventional bipolar (Cros et al., 1992; Cho et al., 2001) stimulating technique at EP such that the anode was positioned over the scapula, while the cathode remained in the conventional supraclavicular location (Roth and Magistris, 1987). In peripheral nerves, the CMAP onset latency is determined by large-diameter, fast-conducting motor bers. We noted that the CMAP onset latency strongly correlated with subjects height for each nerve studied. Such a correlation has only been previously reported for recordings over the ADM muscle with magnetic CRS (Cros et al., 1990). The clinical implication of this nding is that onset latencies should be correlated with patients height, as is routinely done for F response latencies. The site of excitation of the motor axons with CRS relies on three assumptions. Firstly, that the minimum F-wave latency corresponds to conduction in the fastest motor bers (Shahani et al., 1987). Secondly, that the F-wave turnaround time is 1 ms. Thirdly, that the conduction velocity in the proximal segments of the peripheral neuraxis is 60 m/s (Mills and Murray, 1986). The present study has established that the site of excitation of the ulnar nerve was on average 7.4 mm away from the cell body, while that of the median nerve was 14.4 mm away. This is in agreement with our previous ndings for the ulnar nerve (Cros et al., 1990), and closer to the cell body than reported by using the percutaneous electrical stimulation method (Mills and Murray, 1986). However, the site of median nerve excitation was further away from the anterior horn cell, and this may be explained by the fact that APB is predominantly innervated by T1 nerve roots, which are further away from the stimulating cathode than the C8 nerve roots. One limitation of CRS is the discomfort associated with needle insertion and subsequent electrical stimulation. The insertion of the monopolar needle produces similar discomfort to that of an EMG needle when sampling

cervical paraspinal muscles. Although most of our normal volunteers reported discomfort, the mean VAS was 3.8 and in no case was CRS discontinued, suggesting good tolerability of the study. Moreover, none of the subjects experienced serious adverse reactions such as bleeding, infection, or pneumothorax during or after the procedure indicating that this procedure is safe when performed by experienced electromyographers. However, diaphoresis was noted in 2 subjects and a pre-syncopal episode in one. These adverse events may have been avoided by performing the study in a lateral decubitus position rather than a sitting position. Another potential limitation is that the data may not be widely applicable to an older population since the subjects were young. In conclusion, the present study has established normative data for CRS using the monopolar needle technique. In addition, CRS was shown to be a safe neurophysiologic technique with minimal discomfort. Future studies should assess the utility of CRS in diagnosing demyelinating neuropathies.

References
Barker AT, Freeston IL, Jalinous R, Jarratt JA. Magnetic stimulation of the human brain and peripheral nervous system: an introduction and the results of an initial clinical evaluation. Neurosurgery 1987;20:1009. Berger AR, Busis NA, Logigian EL, Wierzbicka M, Shahani BT. Cervical root stimulation in the diagnosis of radiculopathy. Neurology 1987;37: 32932. Cho SC, Stassijns G, Black K, Cros D. Comparison of bipolar and monopolar Erbs point stimulation. Neurology 2001;56(Suppl. 3:A). Cros D, Chiappa KH, Gominak S, Fang J, Santamaria J, King PJ, Shahani BT. Cervical magnetic stimulation. Neurology 1990;40: 17516. Cros D, Gominak S, Shahani B, Fang J, Day B. Comparison of electric and magnetic coil stimulation in the supraclavicular region. Muscle Nerve 1992;15:58790. Kimura J. Collision technique. Physiologic block of nerve impulses in studies of motor nerve conduction velocity. Neurology 1976;26:6802 [Minneap]. Lange DJ, Trojaborg W, Latov N, Hays AP, Younger DS, Uncini A, Blake DM, Hirano M, Burns SM, Lovelace RE, Rowland LP. Multifocal motor neuropathy with conduction block: is it a distinct clinical entity? Neurology 1992;42:497505. Menkes DL, Hood DC, Ballesteros RA, Williams DA. Root stimulation improves the detection of acquired demyelinating polyneuropathies. Muscle Nerve 1998;21:298308. Mills KR, Murray NM. Electrical stimulation over the human vertebral column: which neural elements are excited? Electroencephalogr Clin Neurophysiol 1986;63:5829. Roth G, Magistris MR. Detection of conduction block by monopolar percutaneous stimulation of the brachial plexus. Electromyogr Clin Neurophysiol 1987;27:4553. Sander HW, Menkes DL, Triggs WJ, Chokroverty S. Cervical root stimulation at C5/6 excites C8/T1 roots and minimizes pneumothorax risk. Muscle Nerve 1999;22:7668. Shahani BT, McLeod WN, Bertics GM. Response latencies as a measure of conduction in large-diameter alpha motor axons. Neurology 1987; 37(Suppl. 1):114 [Abstract].