Should you restrict the hand of able-bodied participants? (Abstract title: Influence of a transradial amputation on neuromuscular control of forearm muscles)

ABSTRACT
Following an upper-limb amputation the muscles and
tendons in the amputation stump are often rearranged by a
surgical procedure. One of the purposes of this rearrangement
is to shape the stump as to optimally support the prosthesis
socket, and to create good control sites for a myoelectric
prosthesis using direct control. For the transradial level the
main wrist flexors and extensors are used for the latter
purpose and the remaining muscles are mainly used for
reshaping the stump. This is an interesting phenomenon from
a motor control perspective and questions arise to how the
control strategy of the neuromotor system changes after
amputation when muscles and other tissues are rearranged
and subsequently degenerate. Moreover, the feedback loop is
heavily altered due to absence of a moving limb. This also
appears to have an effect on the electromyogram (EMG) as
demonstrated in several studies in which motion intent was
classified using features of the EMG measured at the forearm.
When comparing classification accuracy between ablebodied
subjects and amputee subjects the accuracy was lower
for the amputees. However, the relative accuracy between
able-bodied participants and amputees is fairly consistent
among a range of classification algorithms. Therefore, many
studies recruit able-bodied subjects and extrapolate their
results to the amputee population.
In this study we aim to investigate how transradial
amputation influences the EMG in an effort to improve the
transferability of results from able-bodied participants to
amputee users. In our study protocol, we simultaneously
measure the EMG at the forearm of both the unaffected and
the affected side of transradial amputees. Participants will
perform bimanual (phantom) movements in two different
conditions. In the ‘restricted-hand condition’, the hand of the
able side is restricted by a brace so the movement
contractions become isometric. In the ‘free-hand condition’,
the hand of the able side is not restricted. The purpose of
restricting the able hand is to simulate the loss of hand
movements while contracting wrist muscles and determine
how this influences the EMG. We hypothesize that the EMG

measured at the able-side in the ‘restricted-hand condition’ is
more similar to the EMG at the affected side than it is in the
‘free-hand condition’. To quantify this, we use a patternrecognition
algorithm to classify the motion intent from both
sides and analyse the resulting classification clusters using
the separability index, repeatability index and the semiprincipal
axes as described in the literature.