Information Technology Reference
In-Depth Information
motion support, we can imagine that it is unnecessary for its control system to be as
accurate and precise as an industrial robot. Instead, it is desirable that active orthosis
could be flexible and versatile, capable of supporting a wide range of motion tasks.
Many aspects of human motion-control strategies are still being considered:
neuroscience investigations are still ongoing and are being conducted by various means
in human and animal subjects. Many “standard” methods are based on data recording
from human and animal subjects, such as:
the application of mechanical disturbances (impulses, vibrations, etc.) to the
limb during natural movements, and the contemporary observation of the limb
disturbance rejection;
•
•
central nervous system activity recording, through the fMRI, during natural
and prototypical motion tasks;
peripheral and central nervous system monitoring through temporally
implanted electrodes.
•
However, there is an increasing need to test neuroscientific hypotheses on human
motion-control theories by implementing them on a model system that is under the full
control of the experimenter. In this way, the results obtained by means of the
“standard” neuroscience methods can be compared with those obtained from the model
system. While this can be achieved to some extent through numerical simulation, these
results are only as good as the accuracy of the numerical model as conceived by
investigators. As a supporting tool to these mathematical analyses, the implementation
of a certain hypothesis on a real mechanical system can reveal the effects of not
modelled dynamics and provide critical insight into how the human system functions in
a real environment. For this reason, the importance and the employment of robotic
tools for neuroscience investigations is developing as a separate study, leading to what
is called neurorobotics (i.e. the fusion of neuroscience and robotics).
The emerging roadmap, involving actions with increasing level of complexity,
establishing the main priorities in the short, middle and long term, is:
strengthening the
“neurorobotics paradigm” and the development of robotic tools, focusing on the
specific feature of human limbs, such as functionally inspired robots mimicking
specific human features (i.e. dynamic properties, impedance, antagonistic actuation
scheme, kinematic and dynamic redundancy) (2015);
developing high-complexity and
fully bio-inspired robotic limbs (2020);
developing high-complexity humanoids for the
investigation of whole body human motion control theories (2025).
Neural-machine interface
There are several ways of tapping into neural information, ranging in hierarchical
location (cortex, spinal cord, peripheral nerves, and nerve ending at muscles) and
invasiveness (direct electrodes [needles, cuffs in tissue]) or surface electrodes (EMG or
electroencephalography [EEG]). The most advanced technology in clinical practice for
controlling prosthetic hands is based on myoelectric control (Zecca, 2002) and
interesting results have been achieved by several groups extracting motor information.
More recently a new neural machine interfacing technology called targeted muscle
reinnervation (TMR) has been developed that improves control of multifunctional
myoelectric upper-limb prostheses. Recently, several strategies to use invasive and
non-invasive interfaces with the peripheral nervous system (PNS) have been
implemented: PNS invasive interfaces can be used to discriminate different neural
