Integrating control and feedback into a trans-femoral prosthesis

In the day to day life of a trans-femoral amputee (TFA) a stumble or a fall is
not uncommon, mainly due to the lack of feedback and control of the
prosthesis[23, 22]. Proprioception and motor control are essential for all
motor tasks, but that is exactly what a TFA is missing in the prosthetic leg.
The ultimate goal of this project is to allow TFA to control their prosthesis
and to give them feedback, where necessary, during activities.

There are multiple objectives to this study. First, we want to find a way of
controlling the prosthesis using electromyography (EMG) of the muscles left in
the stump. Secondly, we want to find a way to give feedback to the human
during certain *states* of the prosthesis. Finally we want to combine the
control and feedback to form a closed-loop system.

For this study multiple experiments will be performed on healthy subjects
(students), for initial testing, and TFAs, for testing and finalizing the
setup. First, to determine EMG patterns, for the design of an algorithm to
control the prosthesis. Secondly, to find the optimal way of giving feedback to
the TFA (vibrotactile and electrotactile stimulation). Finally, evaluation
experiments in TFAs to test the closed loop system.

For the EMG measurements the subjects will receive EMG electrodes on several
muscles and they will be asked to perform activities of daily life (ADL). For
the feedback experiments the subjects will receive vibrotactile or
electrotactile stimulation during ADL. For healthy subjects either no further
intervention takes place or they will walk with a TFP-simulator designed for
non-amputees (TFP-sim). The amputees will use their own socket to which either
their own prosthesis or another (controllable) prosthesis will be attached. For
the combined experiments both EMG electrodes and vibrotactile and / or
electrotactile stimulation will be applied during ADL.

For the healthy subjects participating in the EMG experiments the burden will
lie in the time it takes to prepare and perform the experiments. Some will have
to learn to walk with the TFP-sim, but fall prevention will be provided, so no
risk of falling is involved. For the feedback experiments the same holds,
together with slight discomfort the feedback may give at some stages during the
tests (i.e. during threshold determination slight pain might be felt), but
there is no risk of skin damage or other injuries. For the TFAs EMG experiments
are similar, they are all used to walking with a prosthesis so no risks are
expected there. When providing them with another prosthesis than their own they
will receive proper training and support where necessary. For the feedback
experiments the burden is similar to that in healthy subjects, or less due to a
better developed feedback method. TFA are needed for testing of the system, no
other (patient) group can replace them, their direct benefit will be negligible
although eventually they might profit from a controlled prosthesis.