PHYSIOLOGICAL
STUDY OF TES BY EXACT ACTIVE-ASSISTIVE SYSTEM
Y.
Muraoka, Y. Tomita, T. Fujiwara*, Y. Masakado**
*
**Department
of Rehabilitation Medicine,
ABSTRACT
We developed a system that gave electrical
stimulation with intensity in proportion to the voluntary EMG through the EMG
recording electrodes, and used it for hemiplegic traning. The training effect was observed in long (a month
of the training) and short (ten minutes after the training) period. We
investigated this training effect of this system physiologically.
We carried out two experiments;
1. Observation of disynaptic Ia inhibition between tibial anterior muscle (TA) and soleus muscle (SOL) in hemiplegic patients before and after therapeutic electrical stimulation (TES) to TA.
2. Observation of the convergence onto spinal Ia inhibitory interneurons from Ia afferent and from fast conducting corticospinal axons using a peroneal nerve electrical stimulation and a magnetic cortial stimulation.
The inhibition from peroneal
nerve to SOL did not change, while the inhibition from tibial
nerve to TA decreased. TES changed the synaptic transmission efficiency of
mutual inhibition. The signals from corticospinal
axons and antagonistic muscle nerve converged onto human spinal Ia inhibitory interneurons.
We suggested that (1) short-period effect after TES was caused by the descending
signals passing the efficiency-changed synapse, (2) the efficiency was also
changed by the signal’s from cortex, and (3) the
reorganization in the brain by the biofeedback training is necessary to obtain
and keep the long period effect.
Keywords: hemiplegic
patients, electrical stimulation, magnetic stimulation, biofeedback training
INTRODUCTION
We have developed a system that gave electrical
stimulation with intensity in proportion to the voluntary EMG. The stimulation
was applied to a muscle with the EMG recording electrodes.[1]
Thus the system can exactly assist and amplify a weak muscle contraction by
adding electrical stimulation (Exact Active-Assistive System).
We used the system for the hemiplegic
training. The training effects of muscle strength, range of motion, and
coordination of ankle joints were observed in short (ten minutes after the
fifteen minutes training) and long period (a month of the training). The
training by the system was more effective than the ordinary TES.[2]
The system seems to have several advantages
superior to the ordinary TES. The system can train patients to transmit a
command signal from cortex to the target muscle by using the EMG feedback. The
exact contraction can be produced from transmitted command, because the stimulation
intensity is proportional to the transmitted signal and the same electrodes are
used for recording and stimulation. The patient can
recognize the transmitted command with muscle contraction. The sensory receptor
such as tendon and muscle spindle can transmit afferent signals generated by
the muscle contraction to the spinal cord and the brain. The patients learn
movements by repeating this process, and the ordinary TES does not have it. In
this paper we suggested the necessity of this process adding to ordinary TES by
two experiments.
METHODS
Experiment
1. Disynaptic Ia inhibition between TA and SOL in five hemiplegic patients before and after TES to TA
[3] (TES: the duration of 0.3 ms, the frequency of 20 Hz, monophasic
rectanglar wave, repeating 5 seconds stimulation and
15 seconds rest for 15 minutes)
H reflex at SOL (HSOL)
and at TA (HTA ) could be induced in all 5 subjects and in 3
subjects, respectively. Excitability of the motorneuron
pool was assessed by the size of the HSOL
and HTA.
Subjects were seated comfortably and paired
electrodes were placed over bellies of the soleus and tibial
anterior muscles for recording EMG and the level of the head of the fibula to
stimulate the common peroneal nerve to elicit HTA.
The posterior tibial nerve was stimulated through a
monopolar stimulus electrode placed in the popliteal fossa to elicit HSOL.
The anode was placed over the anterior part of the patera.
The duration of the test stimulus of both the tibial
and the peroneal nerve was 1 ms. The
test HSOL and HTA
were kept between 20 and 40 %, and between 5 and 10% of M-max, respectively.
Conditioning stimulus with the duration of 1 ms were applied to the peroneal nerve for testing Ia disynaptic inhibition of HSOL or to the tibial nerve for inhibition of HTA, using the same electrodes to elicit test H reflexes. For conditioning stimuli to the peroneal nerve, intensity was always kept just at motor threshold (1.0MT) for TA. For the stimulation to the tibial nerve, intensities around 0.7MT for the HSOL were applied.
Conditioning and test stimulation interval (C-T
interval) were set ranging from 0 to 10 in the peroneal
nerve and ranging from –5 to 5 ms in the tibial
nerve. One round was defined as each once in every C-T interval and
unconditioned stimulus (Htest) have been randomly
generated every 5 seconds, and ten rounds were repeated. The inhibition was
demonstrated by plotting the amplitude of the conditioned H-reflexes (Hcond) as a percentage of Htest.
The size of H-reflex of disynaptic Ia inhibition was denoted HIa. HIa
of post-TES were standardized by the mean of HIa of pre-TES,
and the values of all subjects were summarized. These were compared between
pre-TES and post-TES by paired t-test.
Experiment
2. Convergence onto Ia inhibitory interneuron from corticospinal axons and antagonistic muscle nerve in five
intact man
Before second main experiment two preliminary
experiments were carried out.
(1)The above-mentioned
experiment (the inhibition of HSOL experiment) was
carried out to investigate arrival time to Ia
inhibitory interneuron from the peroneal nerve
stimulus in each subject.
(2)After setting
stimulus and recording electrodes at the same positions with Experiment 1, the
8-shaped coil of a magnetic stimulator was held flat on the scalp. Its optimal
position was determined by slight displacements until motor evoked potential
(MEP) of the highest amplitude was recorded from TA. The stimulus intensity was
tuned at threshold of the MEP. The transcranial
magnetic stimulation (TMS) was performed as a conditioning stimulus in order to
investigate arrival time to Ia inhibitory interneuron
from the cortex in each subject.
C-T intervals were set
ranging from -10 to 10 ms. One round was defined as each once in every C-T
interval and Htest have been randomly generated every
5 seconds, and ten rounds were repeated. The inhibition to HSOL
by TMS was demonstrated by plotting Hcond/Htest.
Their arrival time after antagonistic stimulus
and TMS by the preliminary experiments were set and we observed the inhibition
to HSOL in the conditioning stimulus of three
patterns; (a) electrical stimulation to peroneal
nerve, HIa [n] (n: round number), (b) a magnetic cortical stimulation that
induced motor evoked potential at TA, HTMS
[n], (c) (a) and (b), HIa+TMS[n]. One round was defined as once
in every condition patterns and unconditioned stimulus (Htest[n])
have been randomly generated every 5 seconds, and ten rounds were repeated. The
inhibition to HSOL by conditioning
stimulus of four patterns was demonstrated by plotting Hcond/Htest.
The values of HTMS[n]-HIa+TMS[n]
were standardized those of Htest[n]-HIa[n], and were summarized in all subjects. They were
compared by paired t-test.
RESULTS
Experiment 1 Fig.1 (a) and
(b) show the representative results of a hemiplegic
subject, and (c) shows the summarized result of all subjects. C-T intervals of
the disynaptic Ia
inhibition of HTA and HSOL
were 0 and 3 ms, respectively as shown in Fig1.(a),(b). The inhibition from peroneal nerve to HSOL
did not change and the inhibition from tibial nerve
to HTA decreased by TES as shown in Fig.1 (c).
Experiment.2
Fig. 2 (a)-(d) were the summarized results of all subjects. C-T interval in the
peroneal nerve experiment was 2 ms and the interval
in TMS was –1 ms as shown in Fig2. (a), (b). The HSOL/Htest of TMS in figure 2 (c) was more than 1.0, since not
only IPSP but also EPSP were generated to anterior horn cell by TMS that could
not stimulate selectively. The mean of HSOL
was smallest in the condition of both stimulations Ia+TMS.
The amount of disynaptic Ia
inhibition with the effect of the descending impulse (Ia+TMS
- TMS) was significantly larger than that without the effect of the descending
impulse (Ia-test) as shown in Fig. 2 (d).
DISCUSSION
The strong disynaptic Ia inhibition of HTA from tibial
nerve was observed in hemiplegic subjects before TES
and this inhibition significantly decreased after TES. We suppose that it is an
effect of transmission efficiency change of the mutual inhibitory synapse from
flexors inhibitory interneuron to extensors inhibitory interneuron as shown in
Fig.3. TES from peroneal nerve activated the route A
of Fig. 3 and kept it after TES for a few hours. The synapse was also activated
because it was included the route A. The synapse inhibited the inhibition of TA
from SOL, consequently the inhibition of TA decreased. It is reported that the
SOL spasiticity decreased with TES. We expected the
inhibition from the peroneal nerve to the SOL would
change by TES, but it did not be observed. Other effect should be considered
for finding the mechanism.
The effect of disynaptic Ia inhibition of TA from peroneal
nerve significantly increased and was 
influenced
by TMS. It means the corticospinal axons and the peroneal nerve converged onto Ia
inhibitory interneurons of TA. Thus the impulses from
cortex also pass through the route A and they are especially possible to
activate it. (see Fig.4)
Here we suggest the causes of the
decline of voluntary contraction at TA after stroke and training effect of the exact
active-assistive system. (1) Due to brain stroke, the transmission efficiency
of mutual inhibitory synapse from flexors to extensors decreases, since the
frequency of impulses passing through the route A decreases. (2) TES activates
the synaptic transmission efficiency and keeps it for a few hours after TES.
(3) As long as the efficiency is kept high, the signal from the cortex can
transmit and contract the target (carry over effect). (4) The efficiency can
not keep for long time since the descending impulse frequency is small in
stroke patient. (5) The exact active-assistive system train patients to
transmit a command signal from cortex to the target muscle exactly. The
training (motion learning in brain) can increase the descending impulse
frequency and keep it for long time gradually. Therefore the wrong training
(learning) due to wrong feedback by misplacement of electrodes leads the
situation of illness worse. Therefore, appropriate feedback information is
essential for stroke patient.
We suggest the biofeedback
training adding to TES is necessary to acquire long time effect of TES, and the
exact feedback information is essential for training in the stroke patients.
The exact active-assistive system satisfies these conditions.
REFFERENCES
[1]
Y.Muraoka,
Y.Tomita, et.al, EMG-controlled
hand opening system for paraplegia, 6th Vienna international
workshop on functional electrostimulation, Vienna,
1998.9
[2]
R.Shirakawa, Y.Muraoka, et.al,
The effect of neuromuscular electrical stimulation with hemiplegia
inpatients; comparison of different levels of stimulation, 13th
International Congress of The World Confederation for Physical Therapy,
Yokohama, 1999.5
[3]
Y.Okuma, R.G.Lee, Reciprocal Inhibition in Hemiplegia:
correlation with clinical features and recovery, The Canadian Journal of Neurological
Sciences, vol.23, pp.15-23, 1996.
AUTHOR’S ADDRESS
Yoshihiro Muraoka
Institute of Biomedical Engineering
Department of Applied Physics and Physico-informatics, Keio University
3-14-1, Hiyoshi,
Kohoku-ku, Yokohama, 223-8522 JAPAN
e-mail:mura@thx.appi.keio.ac.jp