Cortical Representations of Attempted Hand Movements in Amputees Using 7 Tesla MRI

People with the *locked-in syndrome* (LIS) lost nearly all their motor control,
as a consequence of which communication is made impossible or is heavily
hindered, and people suffering from it rely on assistive devices for their
communication, such as eye trackers or head switches.

Brain-Computer Interfaces (BCIs) are able to restore communication in LIS
patients by recording brain activity. Often the sensorimotor cortex is targeted
for BCI control. The sensorimotor cortex responds to executed movements, but to
attempted movements as well.

Previous results have shown that it is possible to decode hand gestures as used
in sign language. Decoding these acts of language production can be used to
restore communication with a BCI in an intuitive way. For the development of
BCIs, it is important to know what happens to the cortical representations of
hand gestures after denervation, and whether decoding this information from
brain signals is still possible in this situation.

In this study we would like to investigate the effects of denervation on the
cortical representation of finger articulations. We will assess the
decodability of six different hand gestures from the American Manual Alphabet
in people with an amputated right or left arm using 7 tesla functional MRI. We
will compare the results to the decodability of gestures in a non-amputated
control group.

All subjects will practise the hand gestures at home (10 minutes per day for
six days in total). On the scanning day, they will perform the gestures in the
MRI scanner according to the instructions presented on the screen. During the
scanning session, the subject wears a so-called *data glove* on their left
hand, which will record the finger positions. Statistical analysis and
classification will be done off-line.

There are no known risks associated with fMRI acquisition. The technique does
not require administration of any contrast agent or ionizing radiation. The
Utrecht group has ample experience with fMRI scanning (300 sessions per year on
the 7 tesla MRI scanner). The fMRI procedure is painless. Slight discomfort may
occur due to peripheral nerve stimulation during scanning, or due to lying
still with the head and part of the body confined in a tunnel-like device.

The results of this study are important for the brain-computer interface
research in the UMC Utrecht. Individual subjects in this study are not expected
to have any benefits from the outcome of this study.