Optimizing the intermanual transfer effects after training with a prosthetic simulator

People with an upper extremity amputation often choose to have fitted a
prosthesis to restore the functionality for as best as possible. Nevertheless,
about 30% of upper extremity amputees do not use their prosthesis at all due to
a low degree of functional use (Biddiss & Chau, 2007; Dudkiewicz et al., 2004;
Kyberd et al., 1998; Plettenburg, 2002). The functional use of upper extremity
prostheses is not only determined by its function, the technical possibilities,
but also by its functionality, the way the amputee is able to handle the
prosthesis. In an earlier study of our research group is shown that prosthetic
skills can be improved when using intermanual transfer {{136 Romkema,S. 2012}}.

Intermanual transfer implies that when you learn a motor task with one arm, not
only that arm improves, but also the arm at the other side becomes better in
the specific task (Hicks et al., 1983; Karni et al., 1998; Kumar & Mandal,
2005; Lee et al., 2010; Mier & Petersen, 2006; Pereira et al., 2011). The
untrained side thus benefits from the trained side. The effect of intermanual
transfer is shown to be present in prosthetic use, as well in body-powered
(Weeks et al., 2003) as in myo-electric prosthesis (Romkema et al., 2012). We
showed that after training of the *unaffected* side using the simulator, the
level of skills at the start of the prosthetic use with the *affected* side was
increased. This effect can be useful in rehabilitation after an upper limb
amputation, because the training can be started earlier.

It is found that it is of great importance to start to train in the first month
after the amputation to achieve maximum success in prosthetic use (Atkins,
1992; Dakpa & Heger, 1997; Gaine et al., 1997). But in this period often the
wounds are not healed yet and the prosthesis is not finished. To be able to
start to train within these weeks, in our last study (NL 35268.042.11) we used
a prosthetic simulator on the unaffected limb. A prosthesis simulator is an
upper limb prosthesis that can be applied to a sound arm. With the prosthesis
simulator the effects of a myo-electric prosthesis can be mimicked. In
myo-electric prostheses the hand is opened and closed by a motor that is
activated by electrical signals produced by the muscles. The simulator can be
used in the same way. It is applied over the arm, where the prosthetic hand is
placed in front of the sound hand (see figure 1 of the research protocol).
Therefore the training with the simulator is comparable. With an upper limb
prosthetic simulator training can start with the unaffected hand. Because of an
intermanual transfer effect a higher starting level can be reached at the time
the prosthetic training is started on the amputated side.
From our earlier study we know that after training with the simulator on one
arm, the movement times in the other arm decreased. Though, until know it
remains unclear what this training should be like. What tasks will give the
largest results and what time intervals should be in between the training
sessions?

The objective of this study is to determine the optimal myo-electric prosthesis
training for obtaining the largest transfer effect. Here fore we plan to
execute three experiments.

At first it needs to be revealed which trainingtasks leads to the largest
transfer of learning effects in using prosthetic simulators. We use four
different training groups. Three groups train each one of the skills necessary
to control a prostheses (reaching, grasping and force control). The last group
trains a combination of all three training tests. The findings will be compared
with a control group, that did not follow a training with the prosthetic
simulator, though will execute the tests. Therefore, our first goal is to
analyze the transfer effects of different training tasks in prosthetic
simulator use in healthy adults.

The next step is determining the optimal spacing of training with a
myo-electric prosthetic simulator. The same experiment as described above will
be executed, though with different time intervals between the training
sessions. With this the interval with the best effect can be found. There is a
lot of research done on the intensity of the training programs, though often
this is done within 24 hours. Our study will focus on a rehabilitation-like
setting and will therefore be three days long. With this, only a small part of
the research focuses on motor skills. Therefore our aim is to find out how long
the interval needs to be in myo-electric prosthetic use to reach, until two
weeks after the training, the largest effect.

Finally, it will be revealed if the transfer effects are not only present in
prosthetic simulators but also in real prostheses. For rehabilitation it is of
great importance to find out if the effects found in prosthetic simulators are
also present in prosthetic users. At the moment a study on the intermanual
transfer effects in a small amount of patients using a myo-electric prostheses
(maximal four) takes place NL 35268.042.11). Because in the end our study
focuses on these patients we would like to include them in both the study on
the tasks as well as on the study on the spacing. De first to patients will
follow the training with the tasks that has been shown to obtain the best
results. The second two will follow the same training but then with the optimal
time interval. De results of these patients can afterward be compared with the
patients that did not followed a training and were measured for the earlier
study. This is to burden the patient group as less as possible.

Three experiments, each with their own design are presented (See table 1 and 2
from the research protocol). In all experiments the same tests are used.
The goal of the first experiment is to test with which kind of tasks the
intermanual transfer effects are largest in able-bodied adults using the
simulator. There will be four experimental groups. The participants in these
groups learn to use de simulator on one arm (training arm). The other arm (test
arm) is measured to find out if there is an improvement. Each group will train
one of the skills necessary to control a prosthesis. The first group trains
only the force control, e.g. the control over the amount of activation to
prevent an object to be squeezed too hard or to prevent it from dropping. The
second group trains only the reaching movements to learn to adapt to the
changes in inertia caused by the additional weight. The third group trains
reaching movements without the experience of the extra weight of the prosthesis
simulator. In this manner the coordination of the grasping is trained. The
training program of the last group consists of all the three prosthetic skills.
The found effects will be compared to the sham group (the fifth group) that did
not receive any training with the prosthetic simulator, though only with the
sound hand and to the control group that just executes the tests. The different
training programs all take 20 minutes and are executed on three consecutive
days. In our earlier study we found that transfer effects were clearly visible
after five days of training, we expect to be able to measure differences in
three days, while still mimicking a realistic rehabilitation setting. The
measurements consists of a pretest, posttest and retention test (seven days
after ending the training), to be able to measure whether there were learning
effects and whether these effects remained. All tests consist of the same
tasks; functional, grip force control, reaching and grasping tasks, with which
we measure whether participants execute the tasks faster, improve their
coordination and force control. Half of the participants will train their
dominant hand and half will train their non-dominant hand.
In the second experiment we will try to find out what the effect of spacing of
the training on the transfer effect is. As found in the literature, for
different tasks, the largest effects are found with a period of minimal 24
hours between training sessions. Though, there is not a lot of research done in
motor skills and for periods longer than 24 hours. Apart from that, it is shown
that the optimal effect depends on the nature of the task and the time of the
retention test. For this study we therefore choose to use three different
intervals with a minimum of 24 hours. The first group trains daily, the second
group trains every second day, the third group trains with two and three days
in between the training sessions. The training tasks will be chosen based on
the experiment described above; the tasks with the largest effects will be
used. Apart from the three tests as in the last experiment, there will be extra
retention test. The first retention test takes place on day 10, the second two
weeks after the last training session. This will make it possible to show the
effects of the different intervals until two weeks after the training.
The last experiment is meant to generalize the results to patients with an
upper-limb amputee. For this experiment patients with an amputation that will
get a myo-electric prosthesis for the first time will be included. The design
of the experiment is comparable to the for the experiments described above,
while the most effective design will be used. The pretest will be left out
because this is impossible due to the amputation.

In experiment 1 in total six groups of 12 participants train to use a
prosthetic simulator for 20 min during 5 days. In experiment 1 and 2 in total
four groups of 16 participants train to use a prosthetic simulator for 30 min
during 5 days. In experiment 3, four patients with an amputation train (5 times
20 min) with the prosthetic simulator on the unaffected arm. The prosthetic
simulator mimics the functioning of a real prosthesis but can be worn by
able-bodied participants and at the sound side of an amputee patient. The
prosthesis simulator places a prosthetic hand in front of the sound hand.

Performing tasks with a prosthetic simulator does not have any risks.