Modelling Motor Control Paradigms - Computational Neuroscience: A Comprehensive Approach - page 557

should be flat, but this is not the case, as shown in Figure 17.17; on the contrary, the
SDP is characterized by a regular alternation of peaks and valleys: the peaks corre-
spond to time intervals in which the ankle torque and the associated motor commands
are relatively stable; the valleys correspond to time instants in which the ankle torque
rapidly shifts from one stable value to another. Moreover, since the COM & COP
oscillations are globally in phase, it follows that the sequence of peaks of the SDP
must be phase-locked with the COM oscillation.
The statistical analysis of SDP data in normal subjects and in some pathological
conditions [3] has shown that the peak-to-peak time interval is very stable ( T p =
595 37 ms) and is independent of the pathological condition. On the contrary, the

amplitude of the peaks (which is equivalent to the duration of stationary motor com-
mands) and the amplitude of the COP-shifts from one peak to the next one (which
correspond to the amplitude of the anticipatory motor commands) are dependent
upon the task (e.g., open eyes vs. closed eyes) and the nature of the pathological
condition. To some extent, this type of organisation of the posturographic control
action is functionally equivalent to the saccadic oculomotor system, alternating be-
tween two functional states:

±

1. acquisition and fixation of a posturographic target ;

2. quick saccadic jump to the next target.

In the oculomotor system the target is typically visual, whereas in the postural sta-
bilisation system the target must be placed on an invisible point, a little bit beyond
the expected position of the COM. This requires a rather complex sensory process-
ing task, certainly more complex that in the oculomotor case: its basic elements are a
pair of internal models: (i) a multisensory data fusion process for estimating the posi-
tion of the COM and the direction of incipient fall and (ii) a prediction capability for
compensating the intrinsic sensory delays and producing an appropriate anticipatory
postural saccade. Since the falling time constant is a biomechanical parameter and
the motor controller must be time-locked with the partial falls in order to stop them,
it is obvious that the T p parameter must be stable if the subjects, even if affected by
significant motor impairment as in the Parkinson's disease, are indeed able to stand
without support.
On the contrary, pathological conditions are quite likely to affect the precision
of the posturographic saccades due to a partial disruption of the internal feedfor-
ward/feedback models and thus the resulting sway will be larger and more irregular,
as actually occurs.
In a similar experimental situation, Loram and Lakie [39] used an artificial task
in which the subjects were required to activated their ankle musculature in order to
balance a large inverted pendulum: they could easily learn such unstable task and
quasi-regular sway was observed like that in quiet standing. It was observed that
any resting equilibrium position of the pendulum is unstable and can be hold only
temporarily; movements from a resting equilibrium position to another one could
only be accomplished by a ballistic-like throw and catch pattern of torque which
could be used for controlling both position and sway size.