ELECTROMYOGRAPHIC IDENTIFICATION OF SPINAL OSCILLATOR PATTERNS AND RECOUPLINGS IN A PATIENT WITH INCOMPLETE SPINAL-CORD LESION - OSCILLATORFORMATION TRAINING AS A METHOD TO IMPROVE MOTOR ACTIVITIES
G. Schalow et al., ELECTROMYOGRAPHIC IDENTIFICATION OF SPINAL OSCILLATOR PATTERNS AND RECOUPLINGS IN A PATIENT WITH INCOMPLETE SPINAL-CORD LESION - OSCILLATORFORMATION TRAINING AS A METHOD TO IMPROVE MOTOR ACTIVITIES, General physiology and biophysics, 15, 1996, pp. 121-220
A patient with a strongly lesioned spinal cord, sub C5, relearned runn
ing, besides improving other movements, by an oscillator formation tra
ining (rhythmic, dynamic, stereotyped exercise). After 45 days of jump
ing on a springboard and other rhythm trainings, the patient was able
to run 90 m in 4ls (7.9 km/h) (even 9.3 km/h 3 years after the lesion)
besides marching (5.7 km/h), cycling, playing tennis and skiing. FF-t
ype (alpha(1)) (f = 8.3-11.4 Hz) and FR-type (alpha(2)) (f = 6.7 Hz) m
otor unit firings were identified by electromyography (EMG) with surfa
ce electrodes by their oscillatory firing patterns in this patient. In
EMG literature, the alpha(2)-oscillatory firing is called ''myokymic
discharging'' Alternating long and short oscillation periods were meas
ured in FF-type motor units, with changing focus (change from long/sho
rt to short/long oscillation periods). The alternating mean period dur
ations differred by approximately 10 ms. Transient synchronization of
oscillatory firing FF-type motor units was observed with up to two pha
se relations per oscillation cycle. In recumbent position, the phase c
hange in synchronization of two oscillatory firing motor units in the
soleus muscle of one leg correlated with the change from alternating t
o symmetrical oscillatory firing of a third motor unit in the soleus m
uscle of the other leg. This measurement indicates that the alternatin
g oscillatory firing of premotor neuronal networks is correlated with
synchronization of oscillatory firing neuronal subnetworks, i.e. with
coupling changes of oscillators, and is not due to reciprocal inhibiti
on of half-centre oscillators as suggested by the change from alternat
ing to symmetrical oscillatory firing. Coupling changes of oscillatory
firing subnetworks to generate macroscopic (integrative) network func
tions are therefore a general organization form of the central nervous
system (CNS), and are not related to rhythmic movements like walking
or running only. It is proposed that synchronization of spinal oscilla
tors, phase changes in synchronization, changes from alternating to sy
mmetrical firing and backwards, and changes in the focus of alternatin
g oscillatory firing are, among others, physiologic coupling rules of
the human CNS to generate, by ongoing coupling changes of oscillatory
firing subnetworks, integrative functions such as rhythmic and non-rhy
thmic movements. One phase relation between two oscillatory firing alp
ha(1)-motor units was preserved from one volitional leg muscle activat
ion (isometric contraction) to the subsequent one. Since running times
improved upon successive runs for 90 m, the spinal cord seems to be a
ble to store pattern organization for seconds up to minutes. Controlle
d and uncontrolled oscillatory firing of alpha(1)-motor units in volit
ionally activated leg muscles were observed in this patient, which ind
icated that there still were pathologic recruitments of subnetworks af
ter re-learning running and other movements. During walking, running,
and jumping on a springboard, the activation patterns of the vastus la
teralis, hamstrings, tibialis anterior, peronaeus longus, peronaeus br
evis and soleus muscles were recorded (surface electromyography) to be
still pathologic in accordance with partly still pathologic joint rot
ation angles measured kinematically. Especially upon running, the left
knee joint flexion was reduced in swine by a rather permanent activit
y of the rectus femoris combined with an extra burst of the vastus lat
eralis in mid-swing. The recorded abnormalities are due to modificatio
n of the motor program rather than to muscle weakness per se. A furthe
r improvement of the movements of the patient seems possible by improv
ing the motor program, i.e. by improving the functioning of the spinal
pattern generators. By comparing the phasic EMG activity upon walking
, running and jumping on a springboard, the motor program turned out t
o be best for jumping and running, at least with respect to the activi
ty of the left peronaeus longus muscle. This indicates that during the
more rhythmic, dynamic, stereotyped movements (like jumping or runnin
g) more physiologic spinal motor programs were activated by the remain
ing supraspinal drive. Jumping on a springboard generated the most phy
siologic movement pattern, probably by ''Mitbewegung'' (co-movement) a
ctivated by the synchronization of both legs once per jumping cycle, w
hich induced stronger synchronization of right and left movement patte
rn generators by a shared afferent input and cycle resettings of oscil
latory firing subnetworks of the left and right pattern generators. It
is proposed that at least partly, the spinal cord generates stereotyp
ed movements by coupling changes of oscillatory firing subnetworks. Th
e main cause for the movement disorders (and spasticity) occurring fol
lowing spinal cord lesion is the pathologic organization of the functi
onally deteriorated neuronal subunits below the lesion (because of non
use as one reason) in combination with a lesion- and degeneration-indu
ced unbalanced supraspinal and afferent drive, respectively for self-o
rganization of spinal networks. On the basis of our successful therapy
trial and measurements on normal, brain-dead, completely (paraplegic)
and incompletely spinal cord lesioned (tetraparetic) individuals, it
is proposed that the training-induced plasticity of the human CNS to r
e-preformate neuronal networks for up to minutes and permanently has b
een underestimated. The oscillator formation training was successful,
since the rhythm training fitted the rhythmic organization of the CNS
(at least with respect to the premotor alpha(1), alpha(2) alpha(3)-osc
illators), as the rhythmic, stereotyped, dynamic leg movements are mai
nly located in the lumbosacral enlargement of the cord, which was not
lesioned in this patient (lesion sub C5), and could be activated by th
e little remaining supraspinal drive, and since co-movement (''Mitbewe
gung'') was induced once per jumping cycle by the simultaneous afferen
t input. This improvement in movements was made possible by re-preform
ation of spinal neuronal networks for improved self-organization by rh
ythmic movement-induced afferent input, and by rebalanced lesion-adapt
ed drive of supraspinal networks (by lesion-adapted reorganization). T
he type-related single motor axon firing patterns were partly verified
by telemetrically obtained rhythmic EMG patterns. The self-organizati
on of spinal oscillators, including coupling rules between o