Biomedical Engineering Reference
In-Depth Information
150
60
*
130
*
*
*
*
110
40
90
70
20
m5f5
m1 f 1
m0 f 1
m0 f 0
m1f0
m5f5
m1f1
m0f1
m0f0
m1f0
Figure 8.4
The RMSE and the maximal ROM achieved by the subjects with different
FES and robot assistive combinations and with different tracking speeds (circles are for
10
◦
/s, deltas are for 20
◦
/s). Differences with statistical statistical significance (P
0.05) were
denoted by “#” with respect to the FES and robot combinations (ANOVA tests) and “*” to
the tracking speed (t-tests).
<
(i.e., 10
◦
/s and 20
◦
/s). In the tracking tasks, EMG signals were collected from
the muscles of FCR, ECR, biceps brachii (BIC), and triceps brachii (TRI). The
tracking performance was evaluated by EMG parameters (i.e., EMG co-contraction
and EMG activation level), root mean squared error (RMSE) between the target
and the actual tracking positions, the maximal range of motion (ROM) during
tracking. The detailed calculation methods for the EMG parameters could be
found in
Chapter 5
. Each tracking task was repeated twice, and the mean value
of an evaluation parameter was taken as an experimental reading for statistical
analysis. During the evaluation, a rest of 2 minutes was taken by a subject to avoid
fatigue; and the sequence of the evaluation trials with different combinations of
assistance was randomly assigned.
After collected the data from the subjects, the evaluation parameters during
the FES-robot assisted tracking were summarized.
Figure 8.4
shows the RMSE and
the maximal ROM of the subjects during the wrist tracking tasks with different FES
and robot assistive combinations and with the two different tracking speeds, i.e.,
10
◦
/s and 20
◦
/s. The RMSE reflected the wrist tracking accuracy of the tested limb.
It was found that with a faster tracking speed (20
◦
/s), the tracking performance
was lowered when no assistance was provided from the FES-robot system (in case
of m0f0). With pure support from the FES part, the tracking RMSE had the largest
mean values for both tracking speeds. The RMSE had the relatively low values in
the assistive patterns of m5f5, m1f1, and m1f0. It suggested that with the support
from the robot part, or the combined supports from both the FES and robot parts
the tracking performance was improved. Different tracking speeds affected the
maximal ROM greatly
(
Figure 8.4
)
. It was observed that in the tracking trials
with 10
◦
/s the maximal ROMs were significantly higher than those with 20
◦
/s
in assistive patterns of m1f1, m0f1, and m0f0. The mean values of the maximal
ROM for m5f5, m1f1, and m1f0 were higher than those for m0f1 and m0f0, in
which the assistive pattern of m1f1 had the highest mean values for both tracking
speeds. The data in RMSE and the maximal ROM implied that with the supports
