A Novel Continuous Intention-Driven Rehabilitation Robot and Its Training Effectiveness - Biomechatronics in Medicine and Healthcare - page 67

were finished in 7 consecutive weeks. For the interactive group in a training
session, each subject was seated with the paretic arm mounted on the robotic
system as illustrated in Fig. 5.1(b) . I MVE and IMVF at 0 ◦ of the wrist angle
were conducted and repeated 3 times before the training. Then, the subject was
required to conduct voluntary wrist tracking tasks. In each session, there were 14
trials, and each trial contained 5 cycles of wrist extension and flexion. During the
tracking, interactive assistive torques (according to Eqs. 1 and 2 ) were provided
by the robotic system in both flexion and extension phases. In order to obtain the
assistance during tracking, the subjects needed to continuously generate voluntary
muscular effort to acquire assistive torque from the robot; otherwise, if no EMG
activity from the target muscle, the robot would not generate any assistance.
Besides the interactive assistive torque, interactive resistive torques, according to
Eq. 3 , were also applied in the tracking trials in a session. In the study, 10%
and 20% of the maximum wrist torques of IMVE/IMVE were selected for the
generation of the interactive resistive torques, which were administrated to all
trials alternatively in each session, together with the interactive assistive torques,
throughout the training course.
For the passive group, the rehabilitation robotic system (CYBEX and NORM,
Computer Sports Medicine, Inc, USA) was used for the training. The standard
setup for wrist extension and flexion of the CYBEX and NORM system was
adopted for the CPM training (2005). In each session, wrist IMVF and IMVE at
the joint angle of 0 ◦ were first conducted and repeated 3 times. Then, there were
14 training trials, and each trial contained 5 cycles of passive wrist extension and
flexion. The range of motion for the wrist joint was set from 45 ◦ to 60 ◦ ,andthe

palm was moving passively with an angular velocity of 10 ◦ /sec in the range.

−

5.3.2 Training Effects

After the 20 sessions of wrist training assisted with continuous intention-driven
robot on the interactive group, voluntary motor functions for the interactive
group on the wrist and elbow joints were improvement as assessed by FMA
shoulder/elbow and wrist/hand scores ( Fig 5.4 ) . The clinical scores suggested that
the interactive wrist training resulted in voluntary motor improvements: not only
at the trained joint, i.e., the wrist, but also at the elbow joint. These observations
were consistent with previous findings that poststroke training on the distal joints
could increase the motor capacity related to the intralimb proximal joint (Krebs
et al. 1998; Cauraugh et al . 2000). Furthermore, the improvement in the FMA
shoulder/elbow scores was maintained in the 3-month follow up test for the
interactive group. However, for the passive group, no significant changes in the
FMA scores after the training. This suggested that the CPM robot-assisted training
did not benefit the voluntary motor recovery, consistent with other reports that
that CPM did not contribute a lot to the improvement in voluntary motor outcome
(Volpe et al . 2004). The CPM treatment only released the muscle spasticity at
the wrist joint after the training indicated by a reduction in the MAS wrist score