- Email: [email protected]
An experimental study for analyzing nerve conduction velocity
Electroeneephalography and clinical Neurophysiology, 1986, 6 3 : 4 8 4 - 4 8 7
484
Elsevier Scientific Publishers Ireland, Ltd.
Short communication AN EXPERIMENTAL
STUDY
FOR
ANALYZING
NERVE
CONDUCTION
VELOCITY
i
T. NAKANISHI, M. TAMAKI, H. MIZUSAWA, T. AKATSUKA and T. K I N O S H I T A Department of Neurology, Institute of Clinical Medicine, and Department of Biomedical Engineering, Institute of Basic Medical Sciences. University of Tsukuba, Ibaraki 305 (Japan) (Accepted for publication: December 9, 1985)
Summary
Conduction velocities of so called A fibers in the bullfrog's sciatic-peroneal nerve were studied by means of a collision neurography in which a submaximal shock to the distal part of the nerve was used to block descending impulses from a supramaximal shock delivered to the proximal two parts of the same nerve respectively. The onset latency of the response to stimulation delivered to the proximal part was almost unchangeable within a certain range of the stimulus intensity of the distal part and fell into 3 classes, and then conduction velocities of so called A fibers were divided into 3 groups. These findings were in good agreement with those obtained by Erlanger and Gasser (1937) using monophasic recording. On the other hand, the distribution of the external diameter of myelinated nerve fibers examined was unimodal. It may depend on the change in threshold due to stimulation through fluid electrodes that the onset latency of the response of a nerve trunk was divided into 3 classes.
Keywords: nerve conduction velocity - collision neurography
In measuring peripheral nerve conduction velocities, the value of only the fastest fibers has been usually estimated clinically. Recently, many attempts have been made to measure the difference in velocity between the fastest and slowest fibers in a nerve trunk or to estimate the nerve fiber conduction velocity distribution in a nerve bundle using several different methods, such as collision technique (Thomas et al. 1959; Hopf 1963; Gilliatt et al. 1976; Leifer et al. 1977) and computer analysis of the compound action potentials (Barker et al. 1979; C u m m i n s et al. 1979a, b). For the computer analysis, however, some assumptions in regard to the quantitative relationship among conduction velocity, single fiber action potential and fiber diameter are necessary, and there has been little agreement about them. It has been suggested that Hopf's technique is the most reliable for measurement of the motor nerve conduction velocity (Rossi et al. 1983). However, there is also a problem about the relationship between conduction velocity and refractory period in this technique as Barker et al. (1979) have pointed out. Using the collision technique with which the refractory period is not concerned, Thomas et al. (1959) and Gilliatt et al. (1976) tried to determine the range of motor nerve conduction velocities by observing single motor unit action potentials recorded with needle electrodes. Gilliatt et al. (1976) hoped that the use of a collision technique combined with coaxial or bipolar needle recording would make it possible to estimate the full range of axonal velocities for motor units. They pointed out that surface recording was unsatisfactory for measuring the velocity of the slow-conducting fibers. But, it is
1This work was supported by a Grant-in-Aid for scientific research from the Ministry of Education of Japan.
practically difficult to estimate the full range of axonal velocities for motor units using coaxal or bipolar needle electrodes. because the needle electrode can sample only a proportion of the motor units in a muscle. Then, we have had a preliminary experiment to analyze nerve conduction velocity using so-called fluid electrodes with a method of collision neurography unrelated to the refractory period.
Methods and Materials The sciatic-peroneal nerves removed from the limbs of 5 bullfrogs were used. Stimulating and recording procedures were performed fundamentally as previously described (Nakanishi 1982, 1983). In brief, the sciatic-peroneal nerve was mounted through the slot of each partition of a plastic box. Each partition and slot were made air-tight with vaseline. Then, each chamber was filled with Ringer's solution. Three pairs of electrodes (SI, $2, $3) were immersed in the adjacent two chambers with a length of 2 cm on one side to stimulate the proximal segment of the nerve respectively, and bipolar electrodes (R) in the two chambers with a length of 1.5 cm on the other side to record action potentials from the distal segment of the nerve (Fig. 1). Three pairs of stimulators were used, and the 3 stimuli were adjusted independently from each other. In this experiment, submaximal stimulation at S1 was arranged to precede supramaximal stimulation at $2 and $3 by 0.5 msec in order to obtain some separation between the two stimuli, because the shortest conduction time between S1 and $2 was 0.7 msec. Submaximal stimulation was also delivered only to the distal electrodes at SI separately. The intensity of submaximal stimu-
0013-4649/86/$03.50 © 1986 Elsevier Scientific Publishers Ireland, Ltd.
COLLISION N E U R O G R A P H Y
485
lation at S1 (Sli) was progressively increased in small steps by a ratio of 3-5% of the intensity ranging from threshold to maximal. These stimuli were applied at a rate of 1/see. The recording electrodes were connected to a differential amplifier with a frequency response of 1.6 Hz-100 kHz (3 dB down), and the output of the amplifier was led to an analogue computer with horizontal resolution of 40/~sec/point with 256 points/channel. Twenty consecutive nerve action potentials were summated, then digitized and stored on disk. An example of the sequence of an experiment was illustrated in Fig. 1. In the top tracing, the nerve fibers were stimulated through $2 both without any preceding distal shock (Sli: 0 V) and with a weak preceding distal shock (Sli: 0.31 V). In the latter case (Sli: 0.31 V), mixed responses were recorded; the first response was that due to $1 (Sli') and the second that due to $2 (S2'/, which was set up by $2 and was not blocked by
antidromic impulses elicited by $1. The middle tracing showed the nerve action potential evoked by stimulation only at S1 with the same intensity of S1 (Sli: 0.31 V). Thus, the only response to $2 ($2') was obtained by subtraction of the response to S1 (Sli') from the mixed responses ($2' + Sli') following two stimulations at $2 and S1 using the computer (Fig. 1, bottom tracing). According to the similar procedures, the response to $3 ($3') which was set up by $3 and was not blocked by antidromic impulses from SI after stimulation with the same intensity of S1 was also recorded. Conduction velocity of a certain part of the nerve fibers between $2 and $3 was then calculated from the difference in latency to the onset of responses to $2 and $3 ($2' and $3') in the usual way. All experiments were carried out at the temperature of the Ringer's solution (22.4-24.5°C) in which the nerve was immersed. For histological examination, after conduction velocity was esumated, the distal segment of the nerve was removed from the plastic box, then fixed with 2.5% glutaraldehyde and postfixed in a isotonic buffered solution of osmium tetroxide. After tissue blocks were embedded in Epoxy resin, cross-sections 1 /~m thick were stained with toluidine blue and then photographed for morphometry of the nerve fibers.
Results
/l/
" ~ ' - -
S11:O.31V
SII:O.31V
II
( $2'+$11')-S Ii'
!I "
$II:031V
l"v ;J
'V' \ ' / i , , 2 j ' , 2
. . . . .
-
I
I
2 msec
Fig. 1. An example of the response to $2 which was set up by $2 and was not blocked by antidromic impulses elicited by S1. (This was observed in no. 5 nerve.) Stimulation at $2 was preceded 0.5 msec by that at $1. For full description see text.
When the stimulus intensity of S1 was increased from 0 to 0.36 V in a test series, the amplitude of the early component of the response to $2 was progressively decreased, but there was little change in the onset latency within that range of the stimulus intensity (Fig. 2, top tracing). While when the stimulus intensity reached 0.41 V the onset latency was substantially prolonged and then almost unchangeable within the next range from 0.41 to 0.60 V (Fig. 2, middle tracing). Similar findings were observed with further increasing the stimulus intensity of $1 (Fig. 2, bottom tracing). As shown in Fig. 2, with increasing the stimulus intensity of S1, the amplitude of the early component of the responses to $2 and $3 was progressively decreased with little change in the onset latency within a certain range of the stimulus intensity, while when the stimulation at $1 reached a certain intensity the onset latency was prolonged and then almost invariable within the next certain range of the stimulus intensity of S1. When the onset latency of the responses to $2 and $3 was analyzed as mentioned above, they were divided into 3 classes. For example, the first class consisted of the latency of the response obtained when the stimulus intensity of $1 was ranged from 0 to 0.36 V (threshold: 0.25 V), the second that obtained by the stimulus intensity ranging from 0.41 to 0.60 V and the third that obtained by the stimulus intensity ranging from 0.95 to 1.09 V as seen in Fig. 2. Conduction velocity between $2 and $3 was thus divided into 3 groups as shown in Fig. 3. The first was 33.0_+4.7 m/see, the second 25.2_+ 1.8 m / s e c and the third 20.7_+4.7 m/see. The number of myelinated fibers in 4 nerves examined ranged from 855 to 1176, the n u m b e r / m m 2 from 7230 to
486
T. N A K A N I S H 1 ET AL.
,s3, t'I tRt
,ll _
_./li/
( 82'~"81r) - 811"
1
~
.
I F
.
.
.
.
.
.
........
.
.
( 82'4S11')-811' ( S3~,811')-I111'
Sl(:OV
81':0 - 0." V( 0,031,034,0.36V)
F Lllr-r
......
I,J ,Aa :::i .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
t - - I 2rn~
Fig. 2. An example of the effect of increasing the stimulus intensity at $1 on the onset latency observed in no. 5 nerve. Responses to 4 - 8 stimuli of graded intensity were superimposed in each tracing. Note that the onset latency of the response to $2 was almost invariable within a certain range of the stimulus intensity of S1. 11,180. The area occupied by myelinated nerve fibers was 57.3-64.7% (mean 60.2%, S.D. 3.9%) of the total endoneurial space. In all nerves, the distribution of external fiber diameters was unimodal but very broad with the peak at 10/.tm as shown in Fig. 4.
Discussion The collision technique used in this study was not related to refractory period of the nerve, because the descending impulses from the proximal part of the nerve at $2 and $3 had started before the antidromic action potential wave front elicited by S1 reached $2 and $3. It was interesting that nerve conduction velocities of the so-called A fibers in the bullfrog's sciatic-peroneal nerve estimated by this collision technique fell into 3 sharply divided groups. For example, if the threshold of the nerve examined in Fig. 2, that is 0.25 V, is designated as 1, these groups were obtained by stimulus intensity of S1 ranging from 1 to 1.4, from 1.6 to 2.4 and from 3.8 to 4.4 respectively.
If#
,
....
Fig. 3. Three groups of nerve conduction velocity of so-called A fibers in the bullfrog's sciatic-peroneal nerve observed in no. 5 nerve. The responses to both $2 and $3 were superimposed in each tracing. Note that the difference in onset latency to the response to $2 and $3 became substantially longer with increasing the stimulus intensity of S1. These findings were in good agreement with those obtained by Erlanger and Gasser (1937) from the bullfrog's sciatic-peroneal nerve using monophasic recordings. They divided the A peak into 3 waves (alpha, beta, gamma) and described that if the threshold to alpha wave is assigned as a value of 1, alpha was obtained by stimulus intensities ranging from 1 to 1.5, beta by those ranging from 1.6 to 2.4 and g a m m a by those ranging from 3.3 to 4.45. Therefore, the experimental evidence obtained in this study suggests that the collision technique with a method
601
t 50 4o 1 == 30t L 20i U_
5
10 Diameter ()Jm)
15
20
Fig. 4. Distribution of diameters of myelinated nerve fibers obtained from no. 5 nerve. Total fiber number was 1176. Note that the distribution of the fiber diameters was unimodal.
487
COLLISION NEUROGRAPHY of 3-point stimulation might be useful for clinical measurement of the nerve conduction velocities even if surface recording is used. (Our data obtained from h u m a n s using this technique will be published in the near future.) A peculiar finding, however, was observed in this experiment. That is, in spite of progressively increasing the stimulus intensity of S1, the onset latency of the responses to $2 and $3 was almost unchangeable within a certain range of the stimulus intensity and fell into 3 classes, then conduction velocity was divided into 3 groups as mentioned above. On the other hand, the distribution of the external diameter of myelinated nerve fibers in the bullfrog's sciatic-peroneal nerves examined was unimodal as shown in Fig. 4. Since it is generally accepted that there is a linear relationship between nerve conduction velocity and diameter of myelinated nerve fibers (Erlanger and Gasser 1937; Hursh 1939; Tasaki et al. 1944; Boyd 1964; T a c k m a n n et al. 1976), it is a problem that when a nerve trunk with a unimodal histological distribution of fiber diameters was stimulated through fluid electrodes conduction velocities were divided into 3 groups. However, according to Tasaki et al. (1944), the threshold strength increased as the diameter of the fiber decreased, whereas the relation between the reciprocal of the threshold strength and the fiber diameter did not seem to be linear. They suggested that the threshold value might be influenced by the a m o u n t of connective tissue around the nerve. When a nerve bundle is stimulated through fluid electrodes, the anatomical position of a nerve fiber within the nerve trunk may be also important in determining its threshold. Changes in threshold due to stimulation through fluid electrodes may play an essential role in dividing nerve conduction velocities into 3 groups. When a h u m a n nerve trunk is stimulated through surface electrode, a similar finding may also be observed. Therefore, further study concerning these problems may be necessary in the future.
Resume Analyse expbrimentale de la vitesse de conduction nerveuse
Les vitesses de conductions des fibres dites de type A dans le nerf sciatique-prronier du crapeau-buffle ont 6t6 6tudires h l'aide d'une neurographie avec collision, dans laquelle un choc supramaximal appliqu6 h la partie distale du nerf a 6t6 utilis6 pour bloquer les influx descendants provenant d ' u n choc supramaximal appliqu6 respectivement aux deux parties proximales du mdme nerf. La latence du ddbut de la rrponse h la stimulation appliqure h la partie proximale 6tait pratiquement non modifiable dans certaines limites d ' i n t e n s i t ~ du stimulus appliqu6 h la partie distale et l'on pouvait distinguer 3 classes. Les vitesses de conduction des fibres dites ' A ' ont donc 6t6 srparres en 3 groupes. Ces rrsultats sont en accord avec ceux obtenus en enregistrement monophasique par Erlanger et Gasser (1937). En revanche, la distribution des diamrtres externes des fibres nerveuses myrlinis~es 6tudires 6tait unimodale. La srparation en 3 classes des latences de drbut de rrponses du tronc nerveux pourrait d~pendre de la modification des seuils due h ce que la stimulation est appliqure par des 61ectrodes fluide.
We thank Mr. K. Kasaki and Miss. H. Takaki for their kind cooperation in the use of the computer and Miss M. Sugawara for her kind clerical assistance.
References
Barker, A.T., Brown, B.H. and Freeston, I.L. Determination of the distribution of conduction velocities in h u m a n nerve trunks. IEEE Trans. biomed. Engng, 1979, BME-26: 76-81. Boyd, J.A. The relation between conduction velocity and diameter for the three groups of efferent fibres in nerves to m a m m a l i a n skeletal muscle. J. Physiol. (Lond.), 1964, 175: 33P-35P. Cummins, K.L., Perkel, D.H. and Dorfman, L.J. Nerve fiber conduction-velocity distributions. I. Estimation based on the single-fiber and compound action potentials. Electroenceph. clin. Neurophysiol., 1979a, 46: 634-646. Cummins, K.L., Dorfman, L.J. and Perkel, D.H. Nerve fiber conducton velocity distributions. II. Estimation based on two compound action potentials. Electroenceph. clin. Neurophysiol., 1979b, 46: 647-658. Erlanger, J. and Gasser, H.S. Electrical Signs of Nervous Activity. University of Pennsylvania Press, Philadelphia, PA, 1937. Gilliatt, R.W., Hopf, H.C., Rudge, P. and Barafiser, M. Axonal velocities of motor units in the hand and foot muscles of the baboon. J. neurol. Sci., 1976, 29: 249-258. Hopf, H.C. Electromyographic study on so-called mononeuritis. Arch. Neurol. (Chic.), 1963, 9: 307-312. Hursh, J. Conduction velocity and diameter of nerve fibers. Amer. J. Physiol., 1939, 127: 131-139. Leifer, L.J., Meyer, M.A., Morf. M. and Petring, B. Nerve-bundle conduction velocity distribution measurement and transfer function analysis. Proc. IEEE, 1977, 65: 747-755. Nakanishi, T. Action potentials recorded by fluid electrodes. Electroenceph. clin. Neurophysiol., 1982, 53: 343-345. Nakanishi, T. Origin of action potential recorded by fluid electrodes. Electroenceph. clin. Neurophysiol., 1983, 55: 114-115. Rossi, B., Sartucci, F., Stefanini, A., Pucci, G. and Bianchi, F. Measurement of motor conduction velocity with Hopf's technique in myotonic dystrophy. J. Neurol. Neurosurg. Psychiat., 1983, 46: 93-95. Tackmann, W., Spalke, H.J. and Oginszus, H.J. Quantitative histometric studies and relation of number and diameter of myelinated fibers to electrophysiological parameters in normal sensory nerves of man. J. Neurol. (Berl.), 1976. 212: 71-84. Tasaki, I., Ishii, K. and Ito, H. On the relation between the conduction-rate, the fibre-diameter and the internodal distance of the medullated nerve fiber. Jap. J. med. Sci. 1II Biophys., 1944, 9: 189-199. Thomas, P.K., Sears, T.A. and Gilliatt, R.W. The range of conduction velocity in normal motor nerve fibers to the small muscles of the hand and foot. J. Neurol. Neurosurg. Psychiat., 1959, 22: 175-181.
484
Elsevier Scientific Publishers Ireland, Ltd.
Short communication AN EXPERIMENTAL
STUDY
FOR
ANALYZING
NERVE
CONDUCTION
VELOCITY
i
T. NAKANISHI, M. TAMAKI, H. MIZUSAWA, T. AKATSUKA and T. K I N O S H I T A Department of Neurology, Institute of Clinical Medicine, and Department of Biomedical Engineering, Institute of Basic Medical Sciences. University of Tsukuba, Ibaraki 305 (Japan) (Accepted for publication: December 9, 1985)
Summary
Conduction velocities of so called A fibers in the bullfrog's sciatic-peroneal nerve were studied by means of a collision neurography in which a submaximal shock to the distal part of the nerve was used to block descending impulses from a supramaximal shock delivered to the proximal two parts of the same nerve respectively. The onset latency of the response to stimulation delivered to the proximal part was almost unchangeable within a certain range of the stimulus intensity of the distal part and fell into 3 classes, and then conduction velocities of so called A fibers were divided into 3 groups. These findings were in good agreement with those obtained by Erlanger and Gasser (1937) using monophasic recording. On the other hand, the distribution of the external diameter of myelinated nerve fibers examined was unimodal. It may depend on the change in threshold due to stimulation through fluid electrodes that the onset latency of the response of a nerve trunk was divided into 3 classes.
Keywords: nerve conduction velocity - collision neurography
In measuring peripheral nerve conduction velocities, the value of only the fastest fibers has been usually estimated clinically. Recently, many attempts have been made to measure the difference in velocity between the fastest and slowest fibers in a nerve trunk or to estimate the nerve fiber conduction velocity distribution in a nerve bundle using several different methods, such as collision technique (Thomas et al. 1959; Hopf 1963; Gilliatt et al. 1976; Leifer et al. 1977) and computer analysis of the compound action potentials (Barker et al. 1979; C u m m i n s et al. 1979a, b). For the computer analysis, however, some assumptions in regard to the quantitative relationship among conduction velocity, single fiber action potential and fiber diameter are necessary, and there has been little agreement about them. It has been suggested that Hopf's technique is the most reliable for measurement of the motor nerve conduction velocity (Rossi et al. 1983). However, there is also a problem about the relationship between conduction velocity and refractory period in this technique as Barker et al. (1979) have pointed out. Using the collision technique with which the refractory period is not concerned, Thomas et al. (1959) and Gilliatt et al. (1976) tried to determine the range of motor nerve conduction velocities by observing single motor unit action potentials recorded with needle electrodes. Gilliatt et al. (1976) hoped that the use of a collision technique combined with coaxial or bipolar needle recording would make it possible to estimate the full range of axonal velocities for motor units. They pointed out that surface recording was unsatisfactory for measuring the velocity of the slow-conducting fibers. But, it is
1This work was supported by a Grant-in-Aid for scientific research from the Ministry of Education of Japan.
practically difficult to estimate the full range of axonal velocities for motor units using coaxal or bipolar needle electrodes. because the needle electrode can sample only a proportion of the motor units in a muscle. Then, we have had a preliminary experiment to analyze nerve conduction velocity using so-called fluid electrodes with a method of collision neurography unrelated to the refractory period.
Methods and Materials The sciatic-peroneal nerves removed from the limbs of 5 bullfrogs were used. Stimulating and recording procedures were performed fundamentally as previously described (Nakanishi 1982, 1983). In brief, the sciatic-peroneal nerve was mounted through the slot of each partition of a plastic box. Each partition and slot were made air-tight with vaseline. Then, each chamber was filled with Ringer's solution. Three pairs of electrodes (SI, $2, $3) were immersed in the adjacent two chambers with a length of 2 cm on one side to stimulate the proximal segment of the nerve respectively, and bipolar electrodes (R) in the two chambers with a length of 1.5 cm on the other side to record action potentials from the distal segment of the nerve (Fig. 1). Three pairs of stimulators were used, and the 3 stimuli were adjusted independently from each other. In this experiment, submaximal stimulation at S1 was arranged to precede supramaximal stimulation at $2 and $3 by 0.5 msec in order to obtain some separation between the two stimuli, because the shortest conduction time between S1 and $2 was 0.7 msec. Submaximal stimulation was also delivered only to the distal electrodes at SI separately. The intensity of submaximal stimu-
0013-4649/86/$03.50 © 1986 Elsevier Scientific Publishers Ireland, Ltd.
COLLISION N E U R O G R A P H Y
485
lation at S1 (Sli) was progressively increased in small steps by a ratio of 3-5% of the intensity ranging from threshold to maximal. These stimuli were applied at a rate of 1/see. The recording electrodes were connected to a differential amplifier with a frequency response of 1.6 Hz-100 kHz (3 dB down), and the output of the amplifier was led to an analogue computer with horizontal resolution of 40/~sec/point with 256 points/channel. Twenty consecutive nerve action potentials were summated, then digitized and stored on disk. An example of the sequence of an experiment was illustrated in Fig. 1. In the top tracing, the nerve fibers were stimulated through $2 both without any preceding distal shock (Sli: 0 V) and with a weak preceding distal shock (Sli: 0.31 V). In the latter case (Sli: 0.31 V), mixed responses were recorded; the first response was that due to $1 (Sli') and the second that due to $2 (S2'/, which was set up by $2 and was not blocked by
antidromic impulses elicited by $1. The middle tracing showed the nerve action potential evoked by stimulation only at S1 with the same intensity of S1 (Sli: 0.31 V). Thus, the only response to $2 ($2') was obtained by subtraction of the response to S1 (Sli') from the mixed responses ($2' + Sli') following two stimulations at $2 and S1 using the computer (Fig. 1, bottom tracing). According to the similar procedures, the response to $3 ($3') which was set up by $3 and was not blocked by antidromic impulses from SI after stimulation with the same intensity of S1 was also recorded. Conduction velocity of a certain part of the nerve fibers between $2 and $3 was then calculated from the difference in latency to the onset of responses to $2 and $3 ($2' and $3') in the usual way. All experiments were carried out at the temperature of the Ringer's solution (22.4-24.5°C) in which the nerve was immersed. For histological examination, after conduction velocity was esumated, the distal segment of the nerve was removed from the plastic box, then fixed with 2.5% glutaraldehyde and postfixed in a isotonic buffered solution of osmium tetroxide. After tissue blocks were embedded in Epoxy resin, cross-sections 1 /~m thick were stained with toluidine blue and then photographed for morphometry of the nerve fibers.
Results
/l/
" ~ ' - -
S11:O.31V
SII:O.31V
II
( $2'+$11')-S Ii'
!I "
$II:031V
l"v ;J
'V' \ ' / i , , 2 j ' , 2
. . . . .
-
I
I
2 msec
Fig. 1. An example of the response to $2 which was set up by $2 and was not blocked by antidromic impulses elicited by S1. (This was observed in no. 5 nerve.) Stimulation at $2 was preceded 0.5 msec by that at $1. For full description see text.
When the stimulus intensity of S1 was increased from 0 to 0.36 V in a test series, the amplitude of the early component of the response to $2 was progressively decreased, but there was little change in the onset latency within that range of the stimulus intensity (Fig. 2, top tracing). While when the stimulus intensity reached 0.41 V the onset latency was substantially prolonged and then almost unchangeable within the next range from 0.41 to 0.60 V (Fig. 2, middle tracing). Similar findings were observed with further increasing the stimulus intensity of $1 (Fig. 2, bottom tracing). As shown in Fig. 2, with increasing the stimulus intensity of S1, the amplitude of the early component of the responses to $2 and $3 was progressively decreased with little change in the onset latency within a certain range of the stimulus intensity, while when the stimulation at $1 reached a certain intensity the onset latency was prolonged and then almost invariable within the next certain range of the stimulus intensity of S1. When the onset latency of the responses to $2 and $3 was analyzed as mentioned above, they were divided into 3 classes. For example, the first class consisted of the latency of the response obtained when the stimulus intensity of $1 was ranged from 0 to 0.36 V (threshold: 0.25 V), the second that obtained by the stimulus intensity ranging from 0.41 to 0.60 V and the third that obtained by the stimulus intensity ranging from 0.95 to 1.09 V as seen in Fig. 2. Conduction velocity between $2 and $3 was thus divided into 3 groups as shown in Fig. 3. The first was 33.0_+4.7 m/see, the second 25.2_+ 1.8 m / s e c and the third 20.7_+4.7 m/see. The number of myelinated fibers in 4 nerves examined ranged from 855 to 1176, the n u m b e r / m m 2 from 7230 to
486
T. N A K A N I S H 1 ET AL.
,s3, t'I tRt
,ll _
_./li/
( 82'~"81r) - 811"
1
~
.
I F
.
.
.
.
.
.
........
.
.
( 82'4S11')-811' ( S3~,811')-I111'
Sl(:OV
81':0 - 0." V( 0,031,034,0.36V)
F Lllr-r
......
I,J ,Aa :::i .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
t - - I 2rn~
Fig. 2. An example of the effect of increasing the stimulus intensity at $1 on the onset latency observed in no. 5 nerve. Responses to 4 - 8 stimuli of graded intensity were superimposed in each tracing. Note that the onset latency of the response to $2 was almost invariable within a certain range of the stimulus intensity of S1. 11,180. The area occupied by myelinated nerve fibers was 57.3-64.7% (mean 60.2%, S.D. 3.9%) of the total endoneurial space. In all nerves, the distribution of external fiber diameters was unimodal but very broad with the peak at 10/.tm as shown in Fig. 4.
Discussion The collision technique used in this study was not related to refractory period of the nerve, because the descending impulses from the proximal part of the nerve at $2 and $3 had started before the antidromic action potential wave front elicited by S1 reached $2 and $3. It was interesting that nerve conduction velocities of the so-called A fibers in the bullfrog's sciatic-peroneal nerve estimated by this collision technique fell into 3 sharply divided groups. For example, if the threshold of the nerve examined in Fig. 2, that is 0.25 V, is designated as 1, these groups were obtained by stimulus intensity of S1 ranging from 1 to 1.4, from 1.6 to 2.4 and from 3.8 to 4.4 respectively.
If#
,
....
Fig. 3. Three groups of nerve conduction velocity of so-called A fibers in the bullfrog's sciatic-peroneal nerve observed in no. 5 nerve. The responses to both $2 and $3 were superimposed in each tracing. Note that the difference in onset latency to the response to $2 and $3 became substantially longer with increasing the stimulus intensity of S1. These findings were in good agreement with those obtained by Erlanger and Gasser (1937) from the bullfrog's sciatic-peroneal nerve using monophasic recordings. They divided the A peak into 3 waves (alpha, beta, gamma) and described that if the threshold to alpha wave is assigned as a value of 1, alpha was obtained by stimulus intensities ranging from 1 to 1.5, beta by those ranging from 1.6 to 2.4 and g a m m a by those ranging from 3.3 to 4.45. Therefore, the experimental evidence obtained in this study suggests that the collision technique with a method
601
t 50 4o 1 == 30t L 20i U_
5
10 Diameter ()Jm)
15
20
Fig. 4. Distribution of diameters of myelinated nerve fibers obtained from no. 5 nerve. Total fiber number was 1176. Note that the distribution of the fiber diameters was unimodal.
487
COLLISION NEUROGRAPHY of 3-point stimulation might be useful for clinical measurement of the nerve conduction velocities even if surface recording is used. (Our data obtained from h u m a n s using this technique will be published in the near future.) A peculiar finding, however, was observed in this experiment. That is, in spite of progressively increasing the stimulus intensity of S1, the onset latency of the responses to $2 and $3 was almost unchangeable within a certain range of the stimulus intensity and fell into 3 classes, then conduction velocity was divided into 3 groups as mentioned above. On the other hand, the distribution of the external diameter of myelinated nerve fibers in the bullfrog's sciatic-peroneal nerves examined was unimodal as shown in Fig. 4. Since it is generally accepted that there is a linear relationship between nerve conduction velocity and diameter of myelinated nerve fibers (Erlanger and Gasser 1937; Hursh 1939; Tasaki et al. 1944; Boyd 1964; T a c k m a n n et al. 1976), it is a problem that when a nerve trunk with a unimodal histological distribution of fiber diameters was stimulated through fluid electrodes conduction velocities were divided into 3 groups. However, according to Tasaki et al. (1944), the threshold strength increased as the diameter of the fiber decreased, whereas the relation between the reciprocal of the threshold strength and the fiber diameter did not seem to be linear. They suggested that the threshold value might be influenced by the a m o u n t of connective tissue around the nerve. When a nerve bundle is stimulated through fluid electrodes, the anatomical position of a nerve fiber within the nerve trunk may be also important in determining its threshold. Changes in threshold due to stimulation through fluid electrodes may play an essential role in dividing nerve conduction velocities into 3 groups. When a h u m a n nerve trunk is stimulated through surface electrode, a similar finding may also be observed. Therefore, further study concerning these problems may be necessary in the future.
Resume Analyse expbrimentale de la vitesse de conduction nerveuse
Les vitesses de conductions des fibres dites de type A dans le nerf sciatique-prronier du crapeau-buffle ont 6t6 6tudires h l'aide d'une neurographie avec collision, dans laquelle un choc supramaximal appliqu6 h la partie distale du nerf a 6t6 utilis6 pour bloquer les influx descendants provenant d ' u n choc supramaximal appliqu6 respectivement aux deux parties proximales du mdme nerf. La latence du ddbut de la rrponse h la stimulation appliqure h la partie proximale 6tait pratiquement non modifiable dans certaines limites d ' i n t e n s i t ~ du stimulus appliqu6 h la partie distale et l'on pouvait distinguer 3 classes. Les vitesses de conduction des fibres dites ' A ' ont donc 6t6 srparres en 3 groupes. Ces rrsultats sont en accord avec ceux obtenus en enregistrement monophasique par Erlanger et Gasser (1937). En revanche, la distribution des diamrtres externes des fibres nerveuses myrlinis~es 6tudires 6tait unimodale. La srparation en 3 classes des latences de drbut de rrponses du tronc nerveux pourrait d~pendre de la modification des seuils due h ce que la stimulation est appliqure par des 61ectrodes fluide.
We thank Mr. K. Kasaki and Miss. H. Takaki for their kind cooperation in the use of the computer and Miss M. Sugawara for her kind clerical assistance.
References
Barker, A.T., Brown, B.H. and Freeston, I.L. Determination of the distribution of conduction velocities in h u m a n nerve trunks. IEEE Trans. biomed. Engng, 1979, BME-26: 76-81. Boyd, J.A. The relation between conduction velocity and diameter for the three groups of efferent fibres in nerves to m a m m a l i a n skeletal muscle. J. Physiol. (Lond.), 1964, 175: 33P-35P. Cummins, K.L., Perkel, D.H. and Dorfman, L.J. Nerve fiber conduction-velocity distributions. I. Estimation based on the single-fiber and compound action potentials. Electroenceph. clin. Neurophysiol., 1979a, 46: 634-646. Cummins, K.L., Dorfman, L.J. and Perkel, D.H. Nerve fiber conducton velocity distributions. II. Estimation based on two compound action potentials. Electroenceph. clin. Neurophysiol., 1979b, 46: 647-658. Erlanger, J. and Gasser, H.S. Electrical Signs of Nervous Activity. University of Pennsylvania Press, Philadelphia, PA, 1937. Gilliatt, R.W., Hopf, H.C., Rudge, P. and Barafiser, M. Axonal velocities of motor units in the hand and foot muscles of the baboon. J. neurol. Sci., 1976, 29: 249-258. Hopf, H.C. Electromyographic study on so-called mononeuritis. Arch. Neurol. (Chic.), 1963, 9: 307-312. Hursh, J. Conduction velocity and diameter of nerve fibers. Amer. J. Physiol., 1939, 127: 131-139. Leifer, L.J., Meyer, M.A., Morf. M. and Petring, B. Nerve-bundle conduction velocity distribution measurement and transfer function analysis. Proc. IEEE, 1977, 65: 747-755. Nakanishi, T. Action potentials recorded by fluid electrodes. Electroenceph. clin. Neurophysiol., 1982, 53: 343-345. Nakanishi, T. Origin of action potential recorded by fluid electrodes. Electroenceph. clin. Neurophysiol., 1983, 55: 114-115. Rossi, B., Sartucci, F., Stefanini, A., Pucci, G. and Bianchi, F. Measurement of motor conduction velocity with Hopf's technique in myotonic dystrophy. J. Neurol. Neurosurg. Psychiat., 1983, 46: 93-95. Tackmann, W., Spalke, H.J. and Oginszus, H.J. Quantitative histometric studies and relation of number and diameter of myelinated fibers to electrophysiological parameters in normal sensory nerves of man. J. Neurol. (Berl.), 1976. 212: 71-84. Tasaki, I., Ishii, K. and Ito, H. On the relation between the conduction-rate, the fibre-diameter and the internodal distance of the medullated nerve fiber. Jap. J. med. Sci. 1II Biophys., 1944, 9: 189-199. Thomas, P.K., Sears, T.A. and Gilliatt, R.W. The range of conduction velocity in normal motor nerve fibers to the small muscles of the hand and foot. J. Neurol. Neurosurg. Psychiat., 1959, 22: 175-181.