Biomedical Engineering Reference
In-Depth Information
4.4 PROMISES OF ROBOT-ASSISTED THERAPY OF HAND
FUNCTION
This section presents some of the outcomes of clinical studies performed with
the HapticKnob and the HandCARE, and discusses the potential of these robotic
devices to restore hand function of chronic stroke survivors.
In a first pilot study, four chronic post-stroke subjects, more than 2 years after
stroke (P1-P4, 54-83 years of age, 1 female) participated in an eight-week rehabil-
itation therapy, training a combination of exercises with the HapticKnob and the
HandCARE. All subjects were right-handed and suffered from right hemiparesis.
Participants were eligible for the study if they had a minimum manual muscle
testing (MMT) level of 3 (i.e. able to move the arm against gravity, but not able
to tolerate resistance), no more than level 2 spasticity in hand and arm muscles
measured by the Modified Ashworth Scale (MAS, range [0-5], (Bohannon and
Smith (1987))) and presented no other neurological disorder such as apraxia or
tremor. They were able to move the right arm and hand, but had difficulties in
performing many typical ADL.
Therapy sessions were given twice a week, and consisted of 20 minutes of
training with the HandCARE followed by 20 minutes of training with the Haptic-
Knob. Therapy was tailored to the subject's level of impairment, and consisted of
exercises training sensorimotor function where subjects had to actively perform
grasping movements and forearm pronation/supination with the HapticKnob.
Exercises with the HandCARE focused on fine motor control by training finger
coordination, and generation of independent finger force.
All stroke subjects improved their performance in the different exercises dur-
ing the robot-assisted therapy, which was accompanied by improved hand motor
function and force control.
4.4.1 Improvement in Motor Function
In an exercise to train forearm pronation/supination with the HapticKnob, pa-
tients were asked to reach a target forearm orientation by turning the knob against
a resistive load, simulating the action of turning a door knob. The first movement
was always supination, starting from the rest position of the forearm (full prona-
tion). Once the target was reached, the image on the monitor was reoriented to
require an identical returning movement in pronation for realignment.
Figure 4.5
illustrates the evolution of a pronation movement from beginning to the end of
therapy.
Stroke subjects found supination more difficult than pronation, which can be
explained by abnormally high elbowflexor muscle activity increasing the difficulty
of controlling fine movements in supination. However, significant improvements
were found for the supination movement, which was the major goal of the exercise.
In average, the time required to perform the “twisting” movement significantly de-
creased for both supination
(−
25%
)
and pronation
(−
13%
)
(
Fig. 4.5(C)
)
. However,
